USDA LOCAL AND REGIONAL FOOD AID PROCUREMENT PILOT PROJECT · “USDA Local and Regional Food Aid...

USDA LOCAL AND REGIONAL FOOD AID PROCUREMENT PILOT PROJECT Independent Evaluation Report: Annexes December 2012

USDA LOCAL AND REGIONAL FOOD AID PROCUREMENT PILOT PROJECT

INDEPENDENT EVALUATION REPORT: ANNEXES

Contract No.: D11PC18994 The view, opinions, and/or findings contained in this report are those of the author(s) and should not be construed as an official Government position, policy or decision, unless so designated by other documentation.

LRP Independent Evaluation Report: Annexes i

CONTENTS

ANNEX I. BIBLIOGRAPHY ........................................................................... 2

ANNEX II. SUMMARY OF FINDINGS FOR USDA AND FARM BILL REPORTING REQUIREMENTS, BY PROJECT ........... 8

ANNEX III. IMPACT ON INDIVIDUAL MARKETS ................................. 144

ANNEX IV. EVALUATION SCOPE OF WORK ......................................... 207

ANNEX V. DATA COLLECTION INSTRUMENTS ................................. 223

ANNEX VI. METHODOLOGIES .................................................................. 240

LRP Independent Evaluation Report: Annexes 2

ANNEX I. BIBLIOGRAPHY Abrar, Suleiman. “Estimating Supply Response in the Presence of Technical Inefficiency Using the Profit

Function: an Application to Ethiopian Agriculture.” Leicester, United Kingdom: University of Leicester, 2003.

Aker, Jenny C. “Toward Measuring The Impact of the World Food Program’s Purchase for Progress Initiative.” Washington: Center for Global Development, 2008.

Archambault, S. “Capacity and Organizational Deficiencies in the Effort of Ugandan Farmer Groups to Supply Beans and Maize –The Perspective of A Large Buyer.” Ugandan Journal of Agricultural Sciences. 9: 785-790. Washington, Congressional Hunger Center, 2004.

Barrett, Christopher B. “The United States’ International Food Assistance Programs: Issues and Options for the 2007 Farm Bill.” Agricultural Policy for the 2007 Farm Bill and Beyond. Washington: American Enterprise Institute, 2007.

Barrett, Christopher B. and Daniel G. Maxwell. Food Aid after Fifty Years. New York: Routledge (2005).

Benin, Samuel & Rhona Walusimbi. “Impact of the BIDCO Oil Palm Project on the Existing Oilseed Production and Processing Sector.” Kampala, Uganda: International Food Policy and Research Institue, 2004.

Bishop, Helen and Pierre Traoré. Report on USDA-Funded Purchases 2009 – 2011, WFP Mali Final LRP Project Report. Bamako, Mali: WFP Mali Country Office P4P/Purchase for Progress, 2011. 14.

Coulter, Jonathan. “Local and Regional Procurement of Food Aid in Africa: Impact and Policy Issues.” The Journal of Humanitarian Assistance. Greenwich, United Kingdom: University of Greenwich National Resources Institute, 2007.

Dawes, Michael A. “Potential Local Procurement Strategies for WFP/Zimbabwe in the 2011 Marketing Season.” Harare: Floranature (Pvt) Ltd. On behalf of World Food Programme, 2011.

De Schutter, Olivier. The World Trade Organization and the Post-Global Food Crisis Agenda: Putting Food Security First in the International Trade System- Briefing Note 4. New York: United Nations Human Rights Council, 2011.

Dorosh, Paul A., Simon Dradi, and Steven Haggblade. “Regional Trade and Food Security: Recent Evidence from Zambia.” Food Security in Africa: Market and Trade Policy for Staple Foods in Eastern and Southern Africa. Cheltanham, United Kingdom: Edward Elgar and Food and Agriculture Organization, 2007.

Fossi, Filipo. USDA Local and Regional Procurement Pilot Project, Evaluation Report–Final LRP Project Report World Food Programme/Republic of Congo. Unpublished. Brazzaville, Republic of Congo: WFP/RoC Country Office, 2011.

Haan, Nicholas, Nisar Majid and James Darcy. A Review of Emergency Food Security Assessment Practice in Ethiopia. London: Humanitarian Policy Group of Overseas Development Institute on behalf of World Food Programme, 2006.

Hanrahan, Charles E. “Local and Regional Procurement for U.S. International Emergency Food Aid.” Washington: Congressional Research Service, 2010.

Ho, Melissa D. and Charles E. Hanrahan. “International Food Aid Programs: Background and Issues.” Washington, Congressional Research Service, 2010.


Elasticities Database. Ames, Iowa: Food and Agricultural Policy Research Institute, n.d. Web. 12 Feb 2012. http://www.fapri.iastate.edu/

Kingdom of Cambodia, National Committee for Sub-National Development. Kampong Chhnang Provincial Data Book. Kampong Chhang: Kingdom of Cambodia, 2009. 25-26.

Krell, Karola. Survey of National Policy and Activities Related to Food Safety in Countries Eligible for the Codex Trust Fund. Geneva, Switzerland: Food and Agriculture Organization & World Health Organization, 2006.

Lentz, Erin C. et al. “Final Report: A Multidimensional Analysis of Local and Regional Procurement of U.S. Food Aid,” New York: Cornell University, 2012.

Local and Regional Food Procurement in Uganda: An Analytical Review. Kampala: Sserunukuuma and Associates on behalf of World Food Programme, Strategy Policy and Support Division (PDP) Economic Analysis and Development Policy Unit (PDPE), 2005.

Mercier, Stephanie. “Thinking Ahead: What’s Up with U.S. Food Aid Policy in Advance of the 2012 Farm Bill?” Washington: Partnership to Cut Hunger and Poverty in Africa, 2011.

Morgan, L. EAD Briefing Paper: Pakistan Wheat Policy Issues. Islamabad, Pakistan: Government of Pakistan, Economic Affairs Division, 1989.

Mubarik A. 1988. EAD Special Report No. 11.: Supply Response of Major Crops in Pakistan: a Simultaneous Equation Approach Islamabad, Pakistan: Government of Pakistan, Economic Affairs Division, 1988.

Murshid, K.A.S. and Tuot Sokphally. “Cambodia’s Cross Border Economy: An Exploratory Study.” Phnom Penh: Cambodian Development Resource Institute, April 2005. 23-24.

Mutondo, Joao and Bruno de Araujo, Roger Bourgete, and Guy Mullin. “Market Assessment: Food Support for Protection and Promotion of Lives and Livelihoods of the Most Vulnerable People under USDA Grant. An assessment of USDA-funded WFP commodity purchase in Mozambique.” Unpublished. Mozambique: WFP Mozambique Country Office and Kixiquila Serviços de Pesquisa y Consultoria Ltda, 2011.

Organization for Economic Cooperation and Development. “The Development Effectiveness of Food Aid: Does Tying Matter?” 6 May 2009. Web. 14 July 2011. http://www.oecd.org/document/63/0,3343,en_2649_33721_35440255_1_1_1_1,00.html

Paulsson, Carl. “Emergency Food Assistance to Families Affected by Monsoon Floods in Pakistan: Final LRP Pilot Project Report.” Islamabad: WFP Pakistan Country Office, 2011.

Purchase for Progress (P4P): A Primer. Rome: World Food Programme Purchase for Progress Coordination Unit, 2011.

Rolfing, I. “Day 7 Slides for Methods of Case Study Analysis Course.” Ljubljana, Slovenia: European Consortium for Political Research Summer School in Methods and Techniques, 2009.

Sahley, Caroline et al. The Governance Dimensions of Food Security in Nicaragua. Washington: United States Agency for International Development, 2005.

Samake, Michel and Amadou Boura. Catholic Relief Services Mali Final LRP Project Report. Mali: CRS Mali Country Office, 2011.

Sesay, Baika and Saba Hamilton. “Final Report of Field-Based Project Implementation: Project VOICE/Vouchers Offering Incentives to Communities during Emergency Final LRP Project Report CRS Niger.” Baltimore, MD: Catholic Relief Services, 2011.


Toppen, Joel J. “Should the U.S. End In-Kind Food Aid? Assessing the Case for Cash.” Holland, Michigan: Hope College, 2006.

Tschirley, David and Anne Marie del Castillo. “Local and Regional Food Aid Procurement: An Assessment of Experience in Africa and Elements of Good Donor Practice: Working Paper No. 91,” East Lansing, Michigan: Michigan State University, 2007.

“Transitioning U.S. Food Aid Toward Local and Regional Purchase: Lessons from the European Union Experience.” Washington: Action Aid, 2007.

United States Agency for International Development. “AgCLIR: Tanzania. Commercial Legal and Institutional Reform in Tanzania’s Agricultural Sector." USAID Business Climate Legal and Institutional Reform program publication. May 2010. Web. 19 Jun. 2012. http://egateg.usaid.gov/sites/default/files/Tanzania_AgCLIR.pdf

United States Agency for International Development, Famine Early Warning System Network. “Special Report: Rice Market Outlook and Food Security.” Washington: USAID, 2011.

United States Agency for International Development, Legislative and Public Affairs Office. "USAID EFSP Fact Sheet." USAID Emergency Food Security Program. 12 Mar 2012. Web. 7 Apr. 2012. http://www.usaid.gov/our_work/humanitarian_assistance/ffp/fy11.efsp.ofs.pdf

United States Agency for International Development and United States Department of Agriculture. U.S. International Food Assistance Report 2010. n.d. Web. 14 Apr. 2010. <http://www.usaid.gov/our_work/humanitarian_assistance/ffp/fy2010.ifarreport.pdf>

---. “U.S. International Food Aid Distribution Web Based Supply Chain Management (WBSCM).” Internal database documenting all USG international food aid commodity shipments originating in the United States beginning in April 2011 (and thus, in-kind procurement for the purposes of this report). 1 Apr 2011 through 5 Mar 2012.

---. “U.S. International Food Aid Distribution- Food Aid Request Entry System (FARES)/ Processed Commodities Inventory Management System (PCIMS).” Internal database documenting all USG international food aid commodity shipments originating in the United States prior to April 2011 (and thus in-kind procurement for the purposes of this report). 2006-April 2011.

United States Congress. The Food, Conservation, and Energy Act of 2008. Title III, Section 3206. Pub. L. 110-246. 122 Stat. 1840-1844. 18 June 2008.

United States Dept. of Agriculture, Agricultural Marketing Service. "Country of Origin Labeling (COOL)

Frequently Asked Questions." Country of Origin Labeling. 12 Jan 2009. Web. 22 Apr 2012. <http://www.ams.usda.gov/AMSv1.0/getfile?dDocName=STELPRDC5074846>.

---, Foreign Agricultural Service. Interim Guidelines For the Local and Regional Food Aid Procurement Pilot Project. 21 Sept 2009. Web. 8 Aug 2011. <http://www.fas.usda.gov/excredits/FoodAid/LRP/Interim_PPP_Guidelines.pdf>

---, ---. “USDA Local and Regional Food Aid Procurement Pilot Project Budget: FY 2009-2011 Obligations.” Internal, unpublished document provided to MSI for purposes of this evaluation, 2011.

---, Foreign Agricultural Service, Office of Capacity Building and Development . The Use of Local and Regional Procurement in Meeting the Food Needs of those Affected by Disasters and Food Crises. 15 Jan 2009. Web. 8 Aug 2011. <http://www.fas.usda.gov/info/speeches/cr011509.pdf>

---, Food Safety Inspection Service. “Food Safety Inspection Service Glossary.” 26 Nov 2004. Web. 6 Apr 2012. < http://www.fsis.usda.gov/Help/glossary-a/index.asp>

http://egateg.usaid.gov/sites/default/files/Tanzania_AgCLIR.pdf


USDA Local and Regional Procurement Pilot Project, Bangladesh Local and Regional Procurement Project: Final Report on Project Requirements, Land of Lakes Final LRP Project Report. Unpublished. St. Paul, Minnesota: Land of Lakes, 2011.

USDA Local and Regional Procurement Pilot Project, Evaluation Report Final LRP Project Report Catholic Relief Services/ Guatemala. Unpublished. CRS Country Office, 2011.

USDA Local and Regional Procurement Pilot Project, Evaluation Report Final LRP Project Report Catholic Relief Services/Guatemala: Questions/Further Clarification Items. Unpublished. Guatemala: CRS Country Office, 2011.

USDA Local and Regional Procurement Pilot Project, Evaluation Report Final LRP Project Report Catholic Relief Services/Mali: Questions/Further Clarification Items. Unpublished. Mali: CRS Country Office, 2011.

USDA Local and Regional Procurement Pilot Project, Evaluation Report –Final LRP Project Report Fabretto Children’s Foundation/Nicaragua. Unpublished. Arlington, VA: Fabretto Children’s Foundation, 2011.

USDA Local and Regional Procurement Pilot Project, Evaluation Report –Final LRP Project Report International Relief and Development/Cambodia. Unpublished. Arlington, VA: International Relief and Development, 2011.

USDA Local and Regional Procurement Pilot Project, Evaluation Report –Final LRP Project Report Land O’Lakes/Zambia. Unpublished. St. Paul, Minnesota: Land O’Lakes, 2011.

USDA Local and Regional Procurement Pilot Project, Evaluation Report Final LRP Project Report Land O’Lakes/Zambia: Questions/Further Clarification Items. Unpublished. St. Paul, Minnesota: Land O’Lakes, 2011.

USDA Local and Regional Procurement Pilot Project, Evaluation Spreadsheet –Final LRP Project Report Mercy Corps/Niger. Unpublished. Portland, OR: Mercy Corps, 2011.

USDA Local and Regional Procurement Pilot Project, Evaluation Report–Final LRP Project Report United Methodist Committee on Relief/Zimbabwe. Unpublished. New York, NY: United Methodist Committee on Relief, 2011.

USDA Local and Regional Procurement Pilot Project, Evaluation Report –Final LRP Project Report World Food Programme/Cameroon. Unpublished. Cameroon: WFP Cameroon Country Office, 2011.

USDA Local and Regional Procurement Pilot Project, Evaluation Report –Final LRP Project Report World Food Programme/P4P Malawi. Unpublished. Malawi: WFP Malawi Country Office, 2011.

USDA Local and Regional Procurement Pilot Project, Evaluation Report –Final LRP Project Report World Food Programme/Mozambique. Unpublished. Mozambique: WFP Mozambique Country Office and Research on Poverty Alleviation, 2011.

USDA Local and Regional Procurement Pilot Project, Evaluation Report –Final LRP Project Report World Food Programme/P4P Tanzania. Unpublished. Tanzania: WFP Tanzania Country Office and Research on Poverty Alleviation (REPOA), 2011.

USDA Local and Regional Procurement Pilot Project, Evaluation Report Final LRP Project Report WFP/Republic of Congo: Questions/Further Clarification Items. Unpublished. Brazzaville, Republic of Congo: WFP/RoC Country Office, 2011.

USDA Local and Regional Procurement Pilot Project, Evaluation Report –Final LRP Project Report World Vision/Kenya PEAK (Project for Emergency Assistance in Kenya). Unpublished. Nairobi, Kenya: World Vision/Kenya Country Office, 2011.

USDA Local and Regional Procurement Pilot Project, Evaluation Report –Final LRP Project Report World Vision/Uganda Acholi Voucher Program. Unpublished. Kampala, Uganda: World Vision/Uganda Country Office, 2011.


USDA Local and Regional Procurement Pilot Project, Evaluation Report Emergency Food Assistance to Drought-Affected Populations in Chad: WFP Chad Final LRP Project Report. Unpublished. Chad: WFP Chad Country Office, 2011. 6.

USDA Local and Regional Procurement Pilot Project, Evaluation Report Emergency Food Assistance to Drought-Affected Populations in Chad: WFP Chad Final LRP Project Report, questions/further clarification items. Unpublished. Chad: WFP Chad Country Office, 2011. 6.

United States Dept. of Transportation, Maritime Administration. “Cargo Preference: Laws and

Regulations.” N.d. Web. 9 Apr 2012. <http://www.marad.dot.gov/ships_shipping_landing_page/cargo_preference/cargo_laws_and_regulations/Laws_Regs.htm>

United States Government Accountability Office. “International Food Assistance: Better Nutrition and Quality Control Can Further Improve U.S. Food Aid.” Report to Congressional Requesters, GAO- 11-491, 2011. <http://www.gao.gov/assets/320/318210.pdf>

---. “International Food Assistance: Funding Development Projects through Purchase, Shipment and Sale of U.S. Commodities Is Inefficient and Can Cause Adverse Market Impacts.” Report to

Congressional Requesters, GAO-11-636, 2011. <http://www.gao.gov/assets/330/320013.pdf>

---. “International Food Assistance: Key Issues for Congressional Oversight.” Report to Congressional Committees, GAO-09-977SP, 2009. <http://www.gao.gov/assets/300/296312.pdf>

---. “Local and Regional Procurement Can Enhance the Efficiency of U.S. Food Aid, but Challenges May Constrain Its Implementation” Report to Chairman, Subcommittee on Africa and Global Health, House Committee on Foreign Affairs, GAO- 09-570. May 2009. Web. 12 July 2011. <http://www.gao.gov/assets/300/290226.pdf>

---. “Various Challenges Impede the Efficiency and Effectiveness of U.S. Food Aid.” Report to the Senate Committee on Agriculture, Nutrition, and Forestry, GAO-07-560, 2007. <http://www.gao.gov/products/GAO-07-560>

Upton, Joanna. Local Food for Education: LEAP Burkina Faso LRP Final Project Report. Unpublished. Baltimore: Catholic Relief Services, 2011.

Vink, Nick et al. “Food Aid Procurement in South Africa: An Analytical Review of WFP Activities.” Stellenbosch, South Africa: University of Stellenbosch and University of Pretoria on behalf of World Food Programme, Strategy Policy and Support Division (PSP), Economic Analysis Unit (PSPE), 2005.

Walker, David J., Jonathan Coulter and Rick Hodges. Natural Resources Institute, for Department for International Development (2005). “Policy Implications Arising from the Development Impact of Local and Regional Procurement of Food Aid.” Greenwich, U.K.: University of Greenwich, 2005.

World Bank. Eastern Africa: A study of the regional maize market and marketing costs. Agriculture and Rural Development Unit (AFTAR). Sustainable Development department, Country Department 1, Tanzania, Uganda and Burundi, Africa Region, 2009.

World Food Programme. “Experiences of P4P Capacity Building Efforts and Procurement from Farmer Organizations: Kenya, Tanzania, Burkina Faso, Mali, El Salvador, and Nicaragua. With contributions from Guatemala, Honduras.” N.d. 7 Apr. 2012. http://documents.wfp.org/stellent/groups/public/documents/reports/wfp244737.pdf


---. “Food Quality Control Glossary." Food Quality Control. n.d. Web. 28 Apr. 2012. <http://foodquality.wfp.org/Glossary/tabid/200/Default.aspx?PageContentID=617>

---. “Hunger Glossary.” N.d. Web. 24 Apr. 2012. < http://www.wfp.org/hunger/glossary>

---. “Revolution: From Food Aid to Food Assistance. Innovations in Overcoming Hunger.” Eds. Steven Were Omamo, Ugo Gentilini, Susanna Sandstorm. N.d. Web. 5 Apr. 2012. <http://documents.wfp.org/stellent/groups/public/documents/newsroom/wfp228796.pdf>

World Food Programme, Executive Board Regular Session 20-23 February, 2006. “Food Procurement in Developing Countries.” Policy Issues: Agenda Item 5. Rome: 2006 (WFP/EB.1/2006/5‐C). Web. 21 Dec. 2011. <http://www.wfp.org/sites/default/files/Food%20Procurement%20in%20Developing%20Countries%20-%20(2006).pdf >

World Food Programme. “Baseline Markets Report and Monitoring Strategy”. Malawi Country Office. July 2010.

World Food Programme, Food Aid Information System. “Glossary and Abbreviations.” N.d. Web. 24 Apr 2012. < http://www.wfp.org/fais/quantity-reporting/glossary>

---,---. “Quantities Delivered Report by Delivery Mode and Recipient, U.S. Food Aid Donations 2010.” N.d. Web. 14 Apr 2012. < http://www.wfp.org/fais/reports/quantities-delivered- report/run/year/2010/recipient/All/mode/All/cat/All/donor/United%20States%20of%20America/c ode/All/basis/0/subtotal/0>

World Food Programme and Management Systems International. “Purchase for Progress Monitoring and Evaluation System Design Considerations: Final Report.” Unpublished. 2005.

World Vision. “Hunger Glossary.” About Hunger. N.d. Web. 24 Apr. 2012. <http://www.worldvision.org/content.nsf/learn/hunger-glossary-a-e?open&lpos=ctr_txt_A-D>

Xinshen, D. et al. 2008. “Accelerating Africa’s Food Production in Response to Rising Food Prices: Impacts and Requisite Actions.” Discussion Paper 00825. Washington: International Food Policy Research Institute, 2008.

Yu, B and S. Fan. “Rice production responses in Cambodia: Discussion Paper 00939.” Washington: International Food Policy Research Institute, 2009.

Zhuang, Renan & Philip Abbott. “Price Elasticities of Key Agricultural Commodities in China.” China Economic Review, 18.2. West Lafayette, Indiana: Purdue University, 2007.


ANNEX II. SUMMARY OF FINDINGS FOR USDA AND FARM BILL REPORTING REQUIREMENTS, BY PROJECT

This annex is intended to fulfill P.L. 110-246, 122 STAT. 1843, Section (f), 1A, i and 1B, ii, I. For each Local and Regional Procurement (LRP) project, the evaluation team has assembled a short project description as well as a detailed chart indicating the participant’s efforts to comply with all reporting requirements listed in Section (e), 2B, i. of the Food, Energy and Conservation Act of 2008, often referred to as the 2008 Farm Bill. Most tables and figures are drawn directly or adapted from participant reporting; additional tables were provided by the evaluation team from publicly available data on historic and prevailing price data, host and donor country government interference in markets, and other Farm Bill factors.

The evaluation team conducted a review of all project data to identify anomalous or discrepant data points. When data were internally consistent across a given project report and no anomalous data were found, the evaluation team included the data in the dataset. When data reported by projects included discrepancies or artifacts of concern, the evaluation team communicated with the project participants (Private Voluntary Organizations (PVOs) and World Food Programme (WFP)) to examine those issues and assure data quality. In some cases, the evaluation team has updated those costs and timelines as a result of those discussions, to more faithfully and comprehensively represent the cost and time required to carry out each project. The evaluation team performed the cost-effectiveness and timeliness comparisons based on these revised data, which are excerpted here for each field project.

This annex represents the evaluation team’s effort to faithfully record what was submitted to the U.S. Department of Agriculture in compliance with Section (e), 2B, i. The numbering of the evaluation’s “Summary of Findings” charts mirrors the organization of the appropriate section of the Farm Bill, to facilitate the reader’s search for information on each project.


CONTENTS

BANGLADESH ....................................................................................................................... 10 Land O’Lakes Local Procurement Project, Development .......................................................................... 10 BURKINA FASO ..................................................................................................................... 17 CRS Local Procurement Project, Development ........................................................................................... 17 CAMBODIA ............................................................................................................................ 23 IRD Local Procurement Project, Development ........................................................................................... 23 CAMEROON .......................................................................................................................... 30 WFP Local Procurement Project, Emergency .............................................................................................. 30 CHAD ...................................................................................................................................... 40 WFP Regional Procurement Project, Emergency ........................................................................................ 40 REPUBLIC OF CONGO ....................................................................................................... 45 WFP Regional Procurement Project, Emergency ........................................................................................ 45 GUATEMALA ......................................................................................................................... 54 CRS Local Procurement Project, Emergency ............................................................................................... 54 KENYA ................................................................................................................................... 62 World Vision Local Procurement Project, Emergency ............................................................................... 62 MALAWI ................................................................................................................................. 70 WFP Local Procurement Project, Development .......................................................................................... 70 MALI ....................................................................................................................................... 77 CRS Local Procurement Project, Development ........................................................................................... 77 MALI ....................................................................................................................................... 82 WFP Local Procurement Project, Development .......................................................................................... 82 MOZAMBIQUE ..................................................................................................................... 89 WFP Local and Regional Procurement Project, Emergency ...................................................................... 89 NICARAGUA .......................................................................................................................... 95 Fabretto Children’s Foundation Local Procurement Project, Development ........................................... 95 NIGER ................................................................................................................................... 101 CRS Local Procurement Project, Emergency ............................................................................................. 101 NIGER ................................................................................................................................... 107 Mercy Corps Local Procurement Project, Emergency .............................................................................. 107 PAKISTAN ............................................................................................................................. 114 WFP Local Procurement Project, Emergency ............................................................................................ 114 TANZANIA ........................................................................................................................... 118 WFP Local Procurement Project, Development ........................................................................................ 118 UGANDA ............................................................................................................................... 123 World Vision Local Procurement Project, Development ......................................................................... 123 ZAMBIA ................................................................................................................................. 131 Land O’Lakes Local Procurement Project, Development ........................................................................ 131 ZIMBABWE ........................................................................................................................... 139 United Methodist Committee on Relief Regional Procurement Project, Emergency .......................... 139


BANGLADESH

LAND O’LAKES LOCAL PROCUREMENT PROJECT, DEVELOPMENT

Land O’Lakes used funds from the USDA LRP Project to purchase locally produced cereal bars. Land O’Lakes worked with two processors who made the cereal bars from chickpeas, peanuts, puffed rice, sesame seeds, and other commodities. The cereal bars were procured through a soft tender process from processors in Dhaka, Narayanganj, and Manikganj, who procured the commodity ingredients through their normal competitive tender processes (called hard tender in this report). The cereal bars were procured in multiple stages, with tenders taking place in October 2010 and March 2011, procurement of commodities occurring in November 2010, and March and April 2011, and the participant taking possession of the cereal bars throughout 2011.

Land O’Lakes then distributed the bars to 441 primary school feeding programs in three subregional districts─ Bakshiganj, Jamalpur Sadar, and Sarishabari ─ in the Jamalpur administrative district of Bangladesh. On average, 155 cereal bars were distributed per student (i.e. between 20 and 29 bars per month). The total planned number of beneficiaries was 100,000. The number of students reached in January was slightly less than 70,000 and grew to over 101,000 by March. By the end of the program, an average of 105,767 students received cereal bars each day with a total of 17,186,000 cereal bars, approximately 687.4 MT, procured and distributed.

Reporting Requirement Notes i. For each market & commodity procured, describe: I Prevailing and historic

supply, demand and price movements of the market

Table 11 shows that overall, Bangladeshi rice production volumes have gradually increased year by year since the 2006-2007 growing season. This growth is due to increases in yields and number of hectares cultivated. In addition, each year’s beginning stocks have generally increased over time, which has enhanced overall supply. Bangladesh does import a small portion of its overall rice needs, although these imports have declined during the past five years, both in tonnage and in terms of proportion of total supply. Starting in around January-February 2007, rice prices began a steep climb from $244 per MT to an average price in April 2008 of $457.40 per MT, as reported by the Bangladesh Department of Agriculture Marketing (see Figure 1). This rise was likely due to global prices that increased by the same rate during the same time period. While prices declined once again to a low of $252.80 per MT in August 2009, prices rose once again starting in November - December 2009, and reached a peak of $459.40 per MT by February 2011. Since then, due to the arrival of the new crop in the markets, prices declined to $365.30 per MT in June 2011, before increasing to $387.80 per MT in August 2011. While prices are substantially lower than in January-February 2011, they are still much higher than their most recent low point in August 2009 (Figure 1). Outside of more national-level secondary price data on rice as described above, the evaluation team was able to use weekly market price data collected by the project itself for puffed rice and for several of the other commodities used in the cereal bars, including peanuts, chickpeas, and sesame seeds. These four commodities together

1 In this annex, all tables and figures of price and time data are found following this reporting requirement table, as provided to the evaluation team via final project reports. Note that in some instances, the evaluation team returned to project staff with questions related to the data and figures may have been updated in the internal evaluation dataset accordingly. The evaluation team faithfully reproduces here the final version as it was received in reporting requirements to USDA.


Reporting Requirement Notes made up 57 percent of the ingredients in the cereal bars produced for the project. No national price data or other secondary price data is available for the other three commodities outside of the price monitoring done by Land O’Lakes, in coordination with Cornell University and a subcontractor, Data International. This primary market price data, gleaned from capital city, source and distribution area wholesale and retail traders is rather complex, and considering that no information is available for either (a) national throughput levels of peanuts, chickpeas and sesame seeds, (b) short term price elasticity for these products, or (c) the relative level of market integration for these less commonly produced commodities, market monitoring on the overarching national level proves challenging. The evaluation team could thus only utilize the primary market price data on the basis of price visualizing. For greater detail on market impact for this project, please see Annex 3. It is also important to note that while national price data is available for regular rice, Land O'Lakes monitored weekly market prices for puffed rice, as the cereal bar used puffed rice as one ingredient. The wholesale price for puffed rice at the time of procurement was $723/MT, which was about 40 percent higher than the wholesale price of regular rice. This should be taken into account when reviewing the purchase price against the import parity price. Primary data from procurement markets indicated very modest (less than 10%) rises in the price of puffed rice over the project’s duration, peaking in January 2011 and again in May 2011. Procurement market puffed rice prices were much more stable than prices in capital city markets.2 Primary procurement market data indicate nothing beyond gradual seasonal price rises in May and in August for chickpeas, for both wholesalers and retailers. Much of this is attributable to inflation and general agricultural commodity price rises, as indicated in the Land O’Lakes final report. Similar trends occurred in chickpea capital city markets. The price of peanuts in procurement markets rose just over 40 percent from the December 2010 average price to a high in March 2011, but fell again from April to August 2011, ending August at 14 percent above the December 2010 price. Similar price patterns were recorded for capital city markets. Procurement market prices for sesame seeds peaked around late December 2010 - January 2011, with prices reverting to earlier lower levels as the year progressed. Prices are generally higher in capital city markets, but exhibit similar trends.

I Extent of competition for procurement bids

All procurement contracts received two bids each. A two-step process was used to bid and contract cereal bar supplies through processors. The first round of bidding used a soft tender process among three pre-qualified bidders. Land O’Lakes determined that the winning bidder or bidders would need capacity building support to manufacture the bars, as these were new products to the market in Bangladesh. The second round used negotiated prices from the selected bidders, based on competitive bids (though one manufacturer elected not to bid). This process established “indefinite quantity” style contracts that allowed the placing of delivery orders. Land O’Lakes reported that a significant limiting factor in the bidding process was the

2 Please note, in the extensive price data provided by Land O’Lakes as part of project reporting, each price source was listed as either source markets, user markets, or capital markets. Source markets were in Manikgonj, Narayanjangonj, Natore, and Rajshahi upazilas (subdistricts) where significant quantities of the ingredient commodities are produced. Beneficiary markets were in urban centers in the three upazilas where the project was implemented: Sharishabari, Jamalpur Sadar, and Bakshganj in Jamalpur. Finally, capital markets were located around the capital city Dhaka, which was also a source market, as some commodities were purchased by processors there.


Reporting Requirement Notes fact that the cereal bar product was new, and it was necessary to build the production capacity of the processors in order to ensure a reliable and timely supply of cereal bars to the target schools that met quality standards.

II Impact of the procurement of the eligible commodity on producer and consumer prices in the market

Cornell University, as part of the Learning Alliance, conducted an analysis of price data to assess market impacts and found no significant market impacts from cereal bar ingredient procurement or distribution of cereal bars. Since the quantities of ingredient commodities procured by processors were small compared to annual production, Cornell University determined that it was impossible to attribute any price movements to these procurements. In addition, cereal bar distributions had no discernible impact on distribution markets due to the fact that they were unique in the market at the time of distribution. Beyond the national-level secondary price data on rice, the evaluation team used weekly market price data collected by Land O’Lakes for puffed rice, peanuts, chickpeas, and sesame seeds. This primary market price data, gleaned from capital city, source and distribution area wholesalers and retailers is valuable but insufficient for these less commonly produced commodities. As such, market monitoring on the overarching national level is challenging. The evaluation team therefore utilized the primary market price data on the main ingredients through the method of price visualizing. The evaluation team’s analyses of these data indicate that this project had no likely impact on the prices of the affected commodities. Further details are provided in Annex 3 and the market impact chapter in the main body of the report.

III Each government market interference or other activity of the donor country that might have significantly affected the supply or demand of the eligible commodity in the area at which the local or regional procurement occurred

The Bangladeshi government has policies in place to support irrigation supplies, subsidized fertilizer, and electrical power and it maintains a restrictive trade policy with high tariffs. A rice export ban was enforced throughout 2010 while large rice and wheat import quotas were set during the same year (the year of the LRP program) to meet high domestic demand. Land O’Lakes found no evidence of significant host government interference relating to markets or prices during the project period that would have impacted supplies or prices for the major ingredient commodities used in the cereal bars. In 2010, the U.S. International Food Assistance Report cited nearly $50M in food aid to Bangladesh, with 234,000 MT of commodities (wheat, vegetable oil, peas, rice, and lentils) provided through Title II, the McGovern Dole International Food for Education and Child Nutrition Program, and the USDA LRP project.3 The U.S. Agency for International Development (USAID) programmed to ship 59,100 MT of commodities and under the CCC-funded Food for Progress program, USDA programmed to ship 35,000 MT of wheat to Bangladesh in 2011.4

IV Quantities and types of eligible commodities procured in the market

The project procured 17,186,000 cereal bars, weighing a total of 687.4 MT. Ingredients for the bars included puffed rice, peanuts, chickpeas, sesame seeds, and vitamin fortification.

V Timeframe of each procurement of each eligible commodity

The average contracting time for the four cereal bar ingredient procurements was 71 days; the average delivery time was 15 days; and the overall average was 86 days. Cereal bars were delivered to the participant on a delivery schedule, with multiple shipments per each contract signed.

3 United States Agency for International Development and United States Department of Agriculture. U.S. International Food Assistance Report 2010. N.d. Web. 14 Apr. 2010. <http://www.usaid.gov/our_work/humanitarian_assistance/ffp/fy2010.ifarreport.pdf>. 4 US Food Aid Tables FY 2011, <http://www.fas.usda.gov/excredits/FoodAid/Reports/reports.html>.


Reporting Requirement Notes VI Total cost of

procurement, including storage, handling, transport and administrative costs

The average cost per MT across all four shipments of cereal bars was $1,992.76 with $143.38 per MT for TSH5 costs.

ii. Assess:

I Whether the requirements of this section have been met

Land O’Lakes’ reporting was found to be compliant with requirements. Supply and price information is focused on milled rice as opposed to puffed rice which is one of the four primary ingredients in the cereal bars. No supply or secondary price data were available for the other three commodities and the data are generally not regularly monitored in food security literature or on publicly available databases. The evaluation team thus relied solely on Land O’Lakes’ own market monitoring for the remaining commodities.

II Impact of different methodologies and approaches on:

aa Local and regional agricultural producers, including large and small agricultural producers

Land O’Lakes reported that one supplier (PRAN) purchased some ingredients directly from contract farmers thus demonstrating that it is possible that in source markets contracting farmers did obtain a market for their products. PRAN only produced three million out of the 17 million bars for the project while Olympic, the other supplier which produced most of the bars, purchased its ingredients from traders and wholesalers who had little or no connection to farmers. It is unlikely that farmers attributed significant benefits from cereal bar production since the bars utilized such a small portion of annual agricultural production in Bangladesh. Moreover, as the cereal bars were a new product on the marketplace, their distribution in schools did not negatively affect farmers by crowding them out of the market. Land O’Lakes reported that an effort which allows sufficient time to develop local commodity supplies and to support small producers, such as WFP’s Purchase for Progress, could ensure greater direct benefits to farmers and to processors. The short time frame of the project did not enable Land O’Lakes to build the capacity of ingredient producers to the extent that efficiencies could be incorporated into value chains and markets.

bb Markets The evaluation team’s analyses of existing and available data indicate that this project had no likely impact on the prices of the affected commodities. Please see above and Annex 3, as well as the market impact chapter of the main body of the report, for further detail. Land O’Lakes pointed to a few promising positive impacts that may result from the commercialization of the cereal bar production. Land O’Lakes’ contract manufacturer (Olympic) in Bangladesh reported brisk sales of their commercial cereal bar product “Energy.” So far, Olympic has been pilot-testing sales of more than one million bars amounting to 40 MT to assess performance. Although not fortified, this product offers a caloric supplement and adds diversity to the diet as it contains pulses and legumes that often are not included in the Bangladeshi diet. If sales are strong, additional production is planned. Initially, processors hired a significant number of workers, and much of the work on the cereal bars was done by hand. As time progressed, suppliers purchased new machinery that ensured standardized quality and a more efficient production process,

5 TSH is an abbreviation used in this report for all transport, storage and handling costs involved in project procurements.


Reporting Requirement Notes building the capacity of local manufacturing as well as improving the nutrition levels of children in beneficiary schools.

cc Low-income consumers No impact was reported or found by either Land O’Lakes or the evaluators on low-income consumers, as the local procurement did not affect prices of purchased commodities. The evaluation team found that no significant price spikes occurred due to ingredient purchase or cereal bar distribution, thus no effects on low-income consumers would be expected.

dd Program recipients Land O’Lakes found that more children attended school as a result of the project. The project saw a 27 percent increase in overall school attendance. Land O’Lakes reported that households participating in the project were poor or extremely poor and many were unable to feed their children sufficiently during the day. The cereal bars supplemented the child’s diet in these households, and further incentivized school attendance. Land O’Lakes did report that the project’s short time frame meant that the delivery of cereal bars would terminate several weeks prior to the school children’s exam week, which might have an effect on further attendance and on the students’ ability to focus in school. Land O’Lakes noted that school feeding programs had been going on for quite some time prior to the LRP Project, though most distributed wheat flour biscuits to students as opposed to cereal bars. Land O’Lakes reports that students preferred cereal bars to the earlier biscuits, though the students’ parents had mixed opinions on the subject.

III Time of delivery The average time for the contracting stage for four procurements was 71 days; the average time for the delivery stage was 15 days; and the overall average time for contracting and delivery was 86 days. Cereal bars were delivered according to a set delivery schedule as opposed to on one single delivery date.

iii. Compare:

IV Quality and safety assurances

Land O'Lakes contracted SGS Bangladesh, Ltd., to conduct commodity testing and inspection for the cereal bars upon their arrival in the project warehouses; no quality problems were reported. The Institute of Food Science and Technology, a division of the Bangladeshi government’s Council of Scientific and Industrial Research, also conducted tests. There were no breaks in the pipeline due to lack of processor efficiency or non-delivery of quality products. SGS conducted two different trainings for their two processors; 54 participants from PRAN and Olympic attended. Training sessions focused on Good Manufacturing Practices and also on implementing the Hazard Analysis and Critical Control Point system for each step of the manufacturing process, which emphasized quality control starting with the handling of raw ingredients all the way through the packaging and storage phases of the manufacturing process. Though training on Food Borne Illness was planned separately, it was mainstreamed with the aforementioned trainings based on a training needs assessment. These three trainings not only ensured the high quality of the cereal bar products but also improved local manufacturing capacity to adjust to such standards, making the product more marketable for future sales after the termination of the LRP project. Olympic has already begun to build on the groundwork laid during the project, as described above.


TABLE 1: BANGLADESH MILLED RICE PRODUCTION (1,000 MT): 2006-2011

Commodity Attribute Country 2006/2007 2007/2008 2008/2009 2009/2010 2010/2011 2011/2012 Year Change

Rice, Milled Beginning Stocks (1000 MT)

Bangladesh 441 446 546 1,148 640 1,300 660

Rice, Milled Milled Production (1000 MT)

Bangladesh 29,000 28,800 31,000 31,000 32,900 33,000 100

Rice, Milled Rough Production (1000 MT)

Bangladesh 43,504 43,204 46,505 46,505 49,355 49,505 150

Rice, Milled Milling Rate (.9999) (1000 MT)

Bangladesh 6,666 6,666 6,666 6,666 6,666 6,666 0

Rice, Milled MY Imports (1000 MT) Bangladesh 769 2,047 602 92 1,560 800 -760

Rice, Milled TY Imports (1000 MT) Bangladesh 1,570 1,658 150 660 1,400 750 -650

Rice, Milled Total Supply (1000 MT) Bangladesh 30,210 31,293 32,148 32,240 35,100 35,100 0

Rice, Milled Consumption and Residual (1000 MT)

Bangladesh 29,764 30,747 31,000 31,600 33,800 34,200 400

Rice, Milled Ending Stocks (1000 MT)

Bangladesh 446 546 1,148 640 1,300 900 -400

Rice, Milled Total Distribution (1000 MT)

Bangladesh 30,210 31,293 32,148 32,240 35,100 35,100 0

Source: USDA PSD, in Bangladesh Local and Regional Procurement Project: Final Report on Project Requirements, Land of Lakes Final LRP Project Report


FIGURE 1: RICE PRICES (WHOLESALE) AUGUST 2006 TO AUGUST 2011


BURKINA FASO

CRS LOCAL PROCUREMENT PROJECT, DEVELOPMENT

Through the Local Education Assistance and Procurement (LEAP) project, Catholic Relief Services (CRS) locally procured foods for distribution to 364 schools in the Gnagna and Namentenga provinces of Burkina Faso. The project purchased millet through producer cooperatives in surplus millet production regions, and cowpeas through small farmers’ associations in the distribution region. The project worked with producers directly to ensure quality and support marketing and production capacity. The project purchased 182.3 MT of cowpeas; 728 MT of millet; and 101.3 MT of cooking oil. Commodities were purchased via hard and soft tender as well as through vouchers. Suppliers were small agricultural producers contracted through large regional farmer cooperatives and smaller local farmer associations; vegetable oil was purchased from processing companies in the capital city (Ouagadougou).



CRS field staff reported in the final project report that the price analysis methodology used in the evaluation relied on secondary price data for two of the three commodities procured. These data looked at ten years of cowpea and millet prices in seven Burkina Faso markets, including markets from both the procurement and distribution zones. This information was not included in documents made available to the evaluation team. The evaluation team separately found retail millet prices in the capital city of Ouagadougou (Figure 1) to be extremely volatile, and in a broadly competitive cereals market, the team would expect the price series for markets in Namentenga to follow these prices closely. Figure 2 takes the millet prices from Figure 1, deflates them using the consumer price index for Ouagadougou, and re-expresses them, year by year, to focus on seasonality. The deflated prices vary significantly from year to year in both level and profile. On average, prices peak in September at the end of the “hungry season” as the rains start, and fall through the fourth-quarter harvest (and immediate post-harvest) period. Prices thereafter tend to rise until the next hunger season. No additional historical price information was found for cowpeas or vegetable oil. However, the evaluation team reported that its discussion on millet market conduct and performance is also applicable to the cowpea market because cereals traders also deal in them due to similarities in production zones, seasonality of harvest, storage needs, and consumer demand. The cooking oil market is different, particularly in that most cooking oil is imported and has unique storage needs. Therefore the traders involved rarely overlap with those involved in cereals.


The CRS final project report indicates that cooking oil was purchased through a competitive tender with processing companies. LEAP staff submitted a competitive tender via the newspaper and three companies responded. The number of respondents may have been low due in part to the requirements which included the submission of proof of quality standards and direct delivery to schools. The project final report further explained that LEAP staff identified viable producers’ cooperatives for millet. This was undertaken through a few different avenues. The program managers worked with Afrique Verte, an NGO with experience buying locally and working with farmers’ associations and unions, and


Reporting Requirement Notes acquired contacts with some farmers’ organizations (FOs) as well as information about regional grain auctions. LEAP staff attended three such auctions between November 2010 and January 2011, which provided an opportunity to present the LEAP project and introduce CRS as a potential client and helped identify potential millet suppliers. LEAP and Afrique Verte staff identified four large farmer unions which collectively could supply the quantity of millet required by the program. The LEAP team discussed the terms and conditions with each group and asked them to propose the maximum quantity each would be able to provide and at what price. A second meeting was held with each group during which price was debated and specific terms accorded. LEAP staff worked directly with school communities to select and approve vendors who would be providing cowpeas to parent-teacher associations (PTAs) via voucher purchases. Voucher purchases require suppliers to be located in extremely close proximity to school distribution sites, so LEAP and Afrique Verte staff made an effort to reach out to all possible farmers’ associations close to targeted school districts, via Afrique Verte’s local contacts as well as through farmer association registries in local divisions of the Ministry of Agriculture and Hydrology. All farmers’ associations contacted expressed interest and discussions later began with 22 associations who would be able to provide sufficient supply, which LEAP staff set at one metric ton. After the initial round of meetings, subsequent meetings were held between LEAP staff and farmers’ associations to explain how quality testing would be performed prior to delivery. No formal contracts were signed with the associations, since purchases were made via vouchers, but program staff met with association farmers and representatives frequently to reassure them that the quantities requested would still be needed at the agreed-upon time, to continue building trust between PTAs and associations prior to purchase and delivery.


The evaluation team’s analyses of existing and available data, including primary price data collected by LEAP staff and secondary data collected by Cornell University, indicate that this project had had a possible impact on the prices of the millet and cowpeas, two of the three commodities procured. Please see Annex 3 and the market impact chapter of the report for further detail. The price data available for these commodities consisted of three data points from procurement markets– one week before, the week of, and one week after procurement. Millet prices rose 43 percent in the week before the procurement and 25 percent in the week following the procurement. Cowpea prices rose 2 percent in the week preceding the procurement and 11 percent the week after. Vegetable oil prices remained static across both periods. The procurements took place at the end of Burkina Faso’s cool dry season after small farmers had sold the majority of the production that they considered to be surplus to meet their household consumption needs. The procurement period took place during a season when prices typically rise, but prices do not normally rise to the degree that Table 2.2 reflects. The team cannot discount the possibility that the LRP procurements may have played a role in provoking these price rises. Separately, Cornell University performed an analysis of market impact using a different dataset of millet and cowpea prices over 10 years in seven Burkinabè markets in the project’s procurement and distribution zones. The small procurement volumes prompted the Cornell team to combine them with similar local purchases by WFP and estimate the joint effects of these procurements on


Reporting Requirement Notes market prices. The combined local procurements impacted retail prices of the two commodities in the procurement and distribution zones. Impacts were statistically significant for millet in the distribution markets one month and two months after the distributions. For cowpeas, impacts on procurement zone prices were only statistically significant one month after the procurement. However, this result does not provide evidence that the evaluators can ascribe uniquely to the LRP procurement. In conclusion, the team has mixed evidence about the influence of the CRS Burkina Faso LRP project’s procurement on market prices. Market impact is possible, but has not been demonstrated.


The evaluation team found that the government-run Société nationale de gestion de stock de sécurité alimentaire (SONAGESS) manages a 30,000 MT national food security stock whose warehouses are generally run in accordance with international norms. PVOs and other organizations lease space in these warehouses for their food aid operations. SONAGESS also maintains a market information system, focusing on cereal prices. The team also found that the Comité national de secours d’urgence et de réhabilitation distributes food aid to food insecure zones. The team further found that the Government of Burkina Faso sometimes formally bans cereal exports during periods of food shortages in order to halt further price rises by stemming cereal outflows. Implementation at border posts may be strict, but cargo can circumvent border posts if the incentives are large enough. Sometimes official bans provided justification for payment of larger bribes at border posts to allow for an export cargo to cross. In some cases, customs officers enforce bans for exports to one country, but not to another. To the extent that bans effectively lower prices within Burkina Faso, they limit the incentives to farmers to produce more cereals. Donor government involvement in food assistance in 2010 included approximately 9,000 MT of cereals and FBFs through Title II,6 with 7,400 MT programmed through USAID in 2011.7


The project purchased 157.1 MT of cowpeas from small farmer associations; 628.1 MT of millet from larger regional farmer cooperatives/associations; and 72.24 MT of vegetable oil by competitive tender with processing companies.


Average procurement times were as follows: • Cowpeas from farmer associations took on average a total of 118 days

with 89 days for contracting and 29 for delivery; • Millet from farmer cooperatives/associations took on average a total of

90 days, with 60 days for contracting and 30 for delivery; and • Vegetable oil procured via a competitive tender took on average a total

of 85 days, with 43 days for contracting and 42 for delivery. VI Total cost of procurement,

including storage, handling, transportation and administrative costs

Broken down by procurement approach, average costs in USD per metric ton are as follows: Soft tender: Millet cost $339.27/MT for commodities, plus an additional $80.78/MT for TSH.

6 United States Agency for International Development and United States Department of Agriculture. U.S. International Food Assistance Report 2010. N.d. Web. 14 Apr. 2010. <http://www.usaid.gov/our_work/humanitarian_assistance/ffp/fy2010.ifarreport.pdf>. 7 US Food Aid Tables, FY2011, < http://www.fas.usda.gov/excredits/FoodAid/Reports/reports.html>.


Reporting Requirement Notes Hard tender: Vegetable oil cost $2,065/MT for commodities, plus an additional $47.04/MT for TSH. Voucher: Cowpeas cost $512.12 per MT for commodities, plus an additional $20.02/MT for TSH.

ii. Assess: I Whether the requirements of

this section have been met CRS reporting was found to be compliant with requirements. However, no data were found regarding historical market supply and demand of LRP commodities.

II Impact of different methodologies and approaches on: aa Local and regional

agricultural producers, including large and small agricultural producers

Large agricultural producers were not found to have been affected. Cornell University reported that the average farmer in the LEAP region sold 0.4 MT (or four 100 kilogram bags) of cowpeas total in the 2010-2011 season. The CRS Final Project Report indicated that direct sales to LEAP decreased travel time and distance by statistically significant averages of 52 percent and 91 percent respectively for members of farmers’ associations. CRS also reported that those involved also received a higher sales price during the project in 2011 compared to prices in 2010, and had a more profitable season. These benefits also accrued to members of LEAP associations not directly involved in the purchase, although to a lesser degree. Participation in LEAP led to improved storage practices. In turn, all members of LEAP associations learned about the USDA-required quality standards, specifically those related to insects, moisture content, and foreign matter. Other benefits recognized by farmers included sales process capacity building, building trust within farmers’ associations and outside clients, and community reinforcement through partnerships with local PTAs.

bb Markets The evaluation team’s analyses of existing and available data indicate that this project had a possible impact on cowpea and millet prices, but that the impact is not necessarily attributable solely to LRP. The team determined that LRP did not impact the price of vegetable oil. Please see Annex 3 and the market impact chapter of the report for further detail.

cc Low-income consumers The CRS Final Report indicated that the cowpea quantity purchased was very small relative to national production and therefore did not impact consumer prices.

dd Program recipients The project final report stated that pupils clearly preferred locally procured commodities to the bulgur wheat and lentils received under the similar MYAP Title II Program—particularly the cowpeas that are identical to those grown in the region. The PTAs developed a stronger sense of ownership of the school-feeding program, along with the capacity to manage it effectively. Results strongly suggested that food has an important influence on student attendance. Moreover, the LEAP report indicated that in terms of the hyperlocal cowpea procurement which involved sourcing cowpeas from small farmers’ organizations within the same subregion, many of the small farmers providing cowpeas were also parents of children attending the very school receiving such commodities. This in turn provided dual benefits, in the form of additional nutrition for students as well as supplemental income for parents. LEAP families were all satisfied knowing that the food their children ate at school was provided by local farms. Finally, ‘school mothers’ –or mothers of students attending LEAP schools who alternated lunch cooking duties –also reported that they were highly satisfied with the commodities provided. In particular, school mothers indicated that the


Reporting Requirement Notes high quality bags commodities were delivered in could be re-sold after the food was prepared for children, almost covering the full cost of the school lunch they had prepared.

III Time of delivery The average contracting time for cowpeas purchased from farmers’ associations was 89 days and the average delivery time was 29 days, for a total average time of 118 days. The average contracting time for millet purchased from farmers’ cooperatives/associations was 60 days and the average delivery time was 30 days, for a total average time of 90 days. The average contracting time for competitive tender purchases of vegetable oil was 43 days and the average delivery time was 42 days, for a total average time of 85 days.

iii. Compare: IV Quality and safety assurances The evaluation team found that in Burkina Faso, one millet vendor was rejected

because of aflatoxin content and another cowpea vendor was rejected due to supplying low quality foodstuffs. After initially having planned that its staff would do the sampling and testing in-house, CRS negotiated a package deal with the national laboratory that included sampling and testing. Despite the high cost of government sampling, CRS recognized the value of working with the government lab and the sampling and testing was conducted in a very professional manner. Individual producers brought their commodities to a central location where the lab took samples. The LEAP project was subject to the quality standards specified in its agreement with USDA and those suggested by CRS headquarters. During the testing, a few samples came back non-compliant with government standards. For millet, the sole violation was a sample containing a live insect. The farmers’ cooperative responsible for the violation was instructed to treat and clean their stock. Quality issues related to cowpeas mainly involved the presence of dirt and broken beans. Laboratories confirmed that the product was safe and could be sorted by recipients.


FIGURE 1: NOMINAL RETAIL MILLET PRICES IN OUAGADOUGOU 2002 - 2010

Source: FEWSNET

FIGURE 2: DEFLATED OUAGADOUGOU RETAIL MILLET PRICES, 2002 – 2012; 2000 CFA FRANCS PER KILOGRAM

Source: FEWSNET

0

100

200

300

CFA

franc

s per

kilo

gram

me

0

100

200

300

Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul

2002 2003 2004

2005 2006 2007

2008 2009 mean


CAMBODIA

IRD LOCAL PROCUREMENT PROJECT, DEVELOPMENT

International Relief and Development (IRD) implemented a local procurement project in Kampong Chhnang province. IRD procured rice, canned fish, and vegetable oil via hard tenders and iron-fortified fish sauce (IFFS) via direct purchase to provide students with a weekly rice-based meal at school. IRD also established school gardens at the partner schools, using produce from the gardens to supplement the school feeding program, teach new skills, and create small surpluses for home rations. In addition, IRD carried out the construction of irrigation ponds in drought-prone communities through a voucher for work scheme. When redeeming their food vouchers, pond workers were able to select from a range of commodities that included rice, canned fish, instant noodles, and vegetable oil from multiple approved vendors.

Reporting requirement Notes i. For each market & commodity procured, describe: I Prevailing and historic


Annual rice supply and demand data were available for Cambodia and are presented in Table 1. According to the Agricultural Marketing Office at the Ministry of Agriculture, Forestry and Fisheries (MAFF), Cambodian rice supplies have been more than sufficient to meet domestic demand and Cambodia has even been producing surpluses for export since 1996. Surpluses exist in some regions, despite rice deficit pockets in others. Exports of raw paddy are transported to neighboring countries during the harvest period. On occasion, milled rice is re-imported during lean seasons. Even though surplus levels fell during the 2002 to 2004 period due to flood and drought, significant quantities were available for export over consumption levels from 2006 onwards. Historically, the supply of IFFS has been driven by international development projects, and country-level production has been sufficient to meet the contractual requirements of these projects. Cambodia imports about 12,000 to 25,000 MT per year of vegetable oil and also imports canned fish, with no reliable data on quantity available for either, as the MAFF does not keep these data. Importation of these two commodities occurs on a rather ad-hoc basis, in line with the country’s needs, as opposed to a formal import exchange. In a normal year, rice prices generally increase during the dry season (April-May) and the lean season (August-November) and decrease in the harvest and post-harvest seasons (December-March). Five year trends show a significant increase in the price of rice: in local markets, rice prices rose from 1,500 Riel/Kg in 2007 to 2,281 Riel/Kg in 2011. However, for the last four years, prices have fluctuated within a narrower band, as shown in Table 2 below. Rice prices also vary by class – first, or aromatic, followed by second and third classes, which are more often consumed in areas of rural poverty. As shown in Table 2 and Figure 1 below, wholesale and retail prices of rice decreased slightly in 2009 and 2010, but began to increase again in 2011 as a result of significant flooding in 16 of Cambodia’s 24 provinces. The price of fuel and other agricultural inputs also jumped at the time, compared with previous years. However, the price of rice in 2011 was still lower than the price of rice in 2008.


Reporting requirement Notes I Extent of competition for

procurement bids Under the IRD project, commodities were procured through three different approaches: hard tenders, direct purchases and vouchers. Commodities procured using a hard tender include rice, canned fish and vegetable oil. The LRP project received bids through two separate tenders. In the first round, 12 bids were received with milled rice, canned fish and vegetable oil all receiving four bids each. In the second round, seven bids were received for rice, six for canned fish and three for vegetable oil for a total of 18 bids. Only IFFS was procured through a direct purchase. By definition, direct purchases involve participants initiating a discussion with one supplier to provide the commodities, and no formal tender is issued. A direct purchase was used in this circumstance as only one supplier was able to provide the commodity with the appropriate amount of fortification. Test results of IFFS samples from other suppliers indicated that the fortification levels did not meet national standards. In terms of commodities procured via vouchers, 59 vendors were approved and trained to participate in the voucher program. Laborers participating in the food for work activities could redeem the vouchers with any of these approved vendors for staple food commodities. Rice, instant noodles, canned fish and vegetable oil were available under the voucher program with vouchers denominated in the local currency.


The conclusion based on the evaluation team’s market analysis is that the procurement may have had an impact on the retail price of canned fish, although to confirm or refute this conjecture, the team would have needed to know market throughput of canned fish in Phnom Penh, which was not available. Based on the minimal time series data available for this analysis (i.e., prices for the week before, the week of and the week following procurement only), there is no indication of any other price impacts due to LRP procurements. Retail price time series data was not available for milled rice, thus secondary price data on wholesale prices was used instead. For more information on market price impacts for all commodities procured under this project, please refer also to Annex 3.


Host government interference in market supply and demand was not found. Government institutional support for the export of milled rice is weak and fragmented between the General Department of Customs and Excise of Cambodia, the Ministry of Agriculture, Forestry and Fisheries, several private phyto-sanitary institutions, and the Ministry of Commerce. Port charges and delays in customs clearance, as well as official and informal fees drive transport prices up, which prompts exporters and traders to explore less expensive options, such as export of paddy rice through unofficial and thus unregulated black market cross-border trade. This diminishes the potential added value of rice exports to the national economy, as the government is unable to collect export fees or control the flow of trade through illegal, unofficial channels. In recent years, food imports have been subject to a seven percent tariff, constituting the lowest category of a four-band import tariff system. Cambodia also restricts exports of rice, imposing a 10 percent tax on raw materials and a 5 percent tax on processed products (including milled rice).8 The combination of tariffs and export

8 Murshid, K.A.S. and Tuot Sokphally. “Cambodia’s Cross Border Economy: An Exploratory Study.” Phnom Penh: Cambodian Development

Resource Institute, April 2005. 23-24.


Reporting requirement Notes taxes creates a higher effective rate of protection and a greater incentive to sell domestically because of the higher prices prevailing in the domestic market. Anecdotal evidence suggested that in-kind rice shipments from external food aid programs pushed consumer prices down by 25-30 percent, but since in-kind shipments could not be extracted from market data and local distribution areas could not be determined, this could not be verified. In 2010, over 12,000 MT of commodities (including rice, but also vegetable oil, canned salmon, beans, and Corn-Soy Blend (CSB)) were shipped to Cambodia through the McGovern-Dole International Food for Education and Child Nutrition Program9 for school lunches. No shipments for Title II or Food for Progress were planned.


Two procurements of rice were made through hard tenders, the first of 41.50 MT and the second of 72.50 MT. An additional 225.11 MT of rice was procured via the voucher approach. Two procurements of canned fish were made via hard tender, at 12.50 and 23.60 MT respectively. 3.76 MT of canned fish were procured through vouchers. Two procurements of vegetable oil were obtained via hard tender, at 4.50 and 7.27 MT respectively. 4.38 MT of vegetable oil were obtained via vouchers. Two direct purchases of IFFS were made at 1.95 and 2.00 MT each. 7.5 MT of instant noodles were procured via a voucher approach.


The first hard tender for rice took a total of 34 days to complete, with 22 days for contracting and 12 for delivery. The second took 27 days, with 21 days for contracting and only 6 for delivery. Voucher procurement of rice took 48 days overall, from meeting with potential vendors to the first voucher redemptions by beneficiaries. The time frame for procuring the canned fish was very similar to that of the rice procurements detailed above, with the first and second hard tenders taking 32 days and 28 days respectively for both contracting and delivery, and 60 days overall. Voucher procurement of canned fish took 48 days overall. Vegetable oil procurements mirror canned fish and rice procurements in terms of timeliness with the first hard tender procurement taking 30 days total and the second taking 24 days. The first direct purchase of IFFS took 73 days total, with 57 days devoted to contracting and 16 to delivery. The second direct purchase took significantly less time than the first, with 19 days for contracting and 12 for delivery, for a total of 31 days. Instant noodles procured via vouchers took 48 days.

VI Total cost of procurement, including storage, handling, transportation and

Broken down by procurement approach, average costs in USD per metric ton are as follows: Hard tender: Rice cost $381.99/MT;

9 United States Agency for International Development and United States Department of Agriculture. U.S. International Food Assistance Report 2010. N.d. Web. 14 Apr. 2010. <http://www.usaid.gov/our_work/humanitarian_assistance/ffp/fy2010.ifarreport.pdf>.


Reporting requirement Notes administrative costs Canned fish cost $2,004.47/MT;

Vegetable oil cost $1,421.77/MT; and All TSH costs were $37/MT. Direct purchase: IFFS cost $1,761.52/MT, plus $37.00/MT for TSH. Voucher: Rice cost $460.99/MT; Instant noodles cost $1,636.67/MT; Vegetable oil cost $1,670.84/MT; Canned fish cost $2,527.69/MT; and All TSH costs were $15.05/MT for voucher-based purchases.

ii. Assess: I Whether the requirements

of this section have been met

IRD reporting was found to be compliant with requirements; however, historical market data was available only for rice.



Interview data from the MAFF at the national and provincial levels and from rice millers and vendors indicate that the local procurements were of such small size compared to the overall level of production that there was likely no impact on large agricultural producers. Commodity distribution via the voucher program was determined to have had no discernible effect on producers as the program is run in diffuse markets across Kampong Chhnang province. Interview data from millers and information from provincial level MAFF data collection and analysis staff shows no indication of price rises or drops around procurements, but rather a predictable and season-based price pattern. Further, in accordance with miller interview responses, small producers tend to sell right after harvest. Since direct distribution and voucher procurements took place after this period, most small producers no longer maintained enough stock to sell, and therefore impacts on small producers were not seen.

bb Market Interview data from retail vendors, wholesale vendors (including one non-participating wholesale vendor), rice millers, and MAFF data collection and analysis staff, along with the results of a price data analysis from data collected by IRD, indicate that there were no variations in prices due to the local procurements. According to wholesale vendors and millers, the procurements for direct distribution (i.e., the school feeding activities) represented only a fraction of the overall market throughput (e.g., one day’s milling output from the smallest of the millers queried). Due to the small size of the procurements, and the proportionally small role they played in market activity, prices were not affected by the purchases. The same can be said for the voucher program in which small-scale purchases were made from many vendors over a longer span of time. The sole exception is the narrow possibility of a ten percent price spike, which was detected in the price of canned fish after the first direct distribution procurement. The size of the purchase was 12.5 MT. Given the equation for elasticity in the evaluation team’s market analysis, the local procurement would have to be at least 5


Reporting requirement Notes percent of the total Phnom Penh market throughput to have resulted in a price spike of this magnitude. The monthly market throughput would have to be less than 325.2 MT in the same period for the price rise to be a possible consequence of the local procurement. Data on the Phnom Penh market throughput of canned fish, according to a 2005 MAFF study on legal and illegal fish importation, indicates an estimate of imports of 422 MT per month. For the previous year, the figures from this report show an average of nearly 1,000 MT per month. For both figures, the report indicates that much of the imported canned fish is used by low-income consumers and those located far from the central markets. If the figures hold true in 2010-2011, the LRP procurement was most likely too small to have caused the price spike. However, lacking any recent information on canned fish throughput, the evaluation team determined that it is possible that IRD’s procurement of canned fish contributed to a rise in price, but it cannot be confirmed due to conflicting evidence. IRD signed a memorandum of understanding with, and built the capacity of, 59 local vendors. Through the voucher program, a total of $133,951 worth of food (approximately 241 MT) was distributed via voucher exchange to 1,559 people.

cc Low-income consumers No price impact was found from either direct distributions or vouchers on low-income consumers except for the previously mentioned possibility of a spike in canned fish prices that was short-lived. The evaluation team found no other possible price impacts that would have affected low-income consumers.

dd Program recipients In terms of the impact of various distribution approaches, IRD directly provided rice-based meals once a week for 32,057 school children and teachers at 125 schools, improving their food security and nutritional status. IRD used the produce from the school gardens to supplement the school-feeding program. Over the course of the program, 13,866 students and teachers benefitted from the produce. 1,291.2 kg of produce were used for school feeding, 116.8 kg were distributed to schoolchildren practicing school gardening, and 482 kg were sold to purchase seeds. Through the voucher program, IRD constructed 62 community ponds to improve access to water for gardening, livestock use, and domestic household use via the voucher-based Food for Work program. A total of 1,559 households with 7,030 family members directly benefitted from the food for work activities. Simultaneously, IRD also established and trained 62 pond management committees (PMC) to mobilize labor, collect cash contributions, manage a maintenance fund, and maintain the ponds over the long-term. Forty community ponds were completed and fully functional, with an impact on 1,748 households with 7,958 family members. Voucher distributions provided these individuals with a greater choice of commodities from which to choose as a direct result of their labor. In terms of the impact of the several procurement approaches used, voucher procurements improved the livelihoods of the approved vendors by providing them with a constant stream of consumers over the voucher exchange period. Hard tenders offered an opportunity to familiarize local businesses with commercial contracting processes. The two direct purchases of IFFS allowed IRD to purchase a nutritionally fortified condiment, which in turn increased the nutrients students received through the school feeding program. Only one supplier was able to provide IFFS with the required level of fortification, thus without this procurement approach the added nutrients received via IFFS would not have been possible.

III Time of delivery For direct distributions, procurement times averaged 35 days overall, with 27 days for contracting and 8 days for delivery. For vouchers, the average overall time for procurement was 48 days.


Reporting requirement Notes iii. Compare: Quality and safety

assurances IRD contracted with the Ministry of Agriculture’s CamControl for quality testing of the commodities. This included testing of vendor products for the voucher program. Delivery timelines and quality testing requirements were met for all procurements, although some vendors expressed concerns during site visit interviews. Concerns included the complaint that testing added to the time vendors had to wait to receive payment, and that since testing was done after delivery to the IRD warehouse, suppliers took on a greater risk during that period. No quality issues were identified with the rice, canned fish, or vegetable oil, but none of the suppliers for the first round of procurement met the Cambodian food quality standards/specifications for IFFS. An examination was conducted at a government laboratory to test for protein content in the IFFS. Results indicated that the IFFS from the marketplace was unacceptable due to low levels of amino acids. As a result, IRD decided to work with one supplier who was willing to improve the quality of his IFFS for the second round of procurement. Testing indicated that the resulting IFFS met all standards. IRD then made two direct purchases from that supplier in order to ensure that the protein content of its school feeding stock was adequate.

TABLE 1: RICE SUPPLY AND DEMAND 2003-2007

Description 2003 2004 2005 2006 2007 Cultivated area (ha) 2,314,285 2,374,175 2,443,530 2,541,432 2,584,907 Production (MT) 4,710,957 4,170,284 5,986,179 6,264,030 6,727,138 Food requirement per year (MT) 1,936,565 1,905,896 2,013,533 2,053,983 2,090,000 Surplus of milled rice (MT) 686,496 416,118 1,319,571 1,433,829 1,640,000 Surplus of paddy rice (MT) 1,072,650 650,184 2,061,830 2,240,358 2,570,000

Source: Ministry of Agriculture, Forestry and Fisheries (MAFF), in USDA Local and Regional Procurement Pilot Project, Evaluation Report –Final LRP Project Report International Relief and Development/Cambodia

TABLE 2: AVERAGE MARKET PRICE OF RICE, 2008 TO 2011

Sale in Category 2008 2009 2010 2011

Wholesale, (Riel/Kg) 1st Class 2,911.90 2,720.15 2,603.13 2,750.00 2nd Class 2,769.84 2,253.95 2,108.33 2,281.25 3rd Class 2,147.62 1,859.50 1,800.00 2,018.75

Retail Sale, (Riel/Kg) 1st Class 3,176.19 2,954.55 2,765.97 2,925.00 2nd Class 2,778.57 2,456.15 2,278.13 2,475.00 3rd Class 2,336.90 2,024.29 1,956.67 2,150.00

Source: Agriculture Marketing Office/MAFF, in USDA Local and Regional Procurement Pilot Project, Evaluation Report –Final LRP Project Report International Relief and Development/Cambodia


FIGURE 1: AVERAGE OF RETAIL SALE PRICE FROM 2008 TO 2011, RIEL/KG

Source: USDA Local and Regional Procurement Pilot Project, Evaluation Report –Final LRP Project Report International Relief and Development/Cambodia.

2008 2009 2010 20111st Class 3176.19 2954.55 2765.97 2925.002nd Class 2778.57 2456.15 2278.13 2475.003rd Class 2336.90 2024.29 1956.67 2150.00

0.00

500.00

1000.00

1500.00

2000.00

2500.00

3000.00

3500.00Av

erag

e Pr

ice,

Rie

l/Kg


CAMEROON

WFP LOCAL PROCUREMENT PROJECT, EMERGENCY

The World Food Programme (WFP) project in Cameroon purchased 329 MT of sorghum, 1,790 MT of maize, and 543 MT of beans through large-scale traders who sourced the commodities from local farmers in the northern and western parts of the country. All three procurements were made through hard tenders, with multiple bids received for each tender. The commodities were delivered to WFP’s warehouses on a monthly basis. Commodities were then distributed directly from the warehouses. In the Logone and Chari regions, 216,000 people affected by drought in the extreme north of the country benefitted from the activities of this project.

Reporting Requirement Notes

i. For each market & commodity procured, describe:

I Prevailing and historic supply, demand and price movements of the market

In Cameroon, WFP procured 329 MT of sorghum from the far northern region, 1,790 MT of maize grain in the northern region, and 543 MT of beans in the western region. Domestic cereal market production is highly dependent on rainfall, especially in the arid northern regions. The North and Far North regions of the country produce a large portion of the country’s cereal crop, with two harvests obtained over the course of the year in September-October and December-February, according to WFP’s final project report. These regions produce approximately 300,000 MT of maize grain and 600,000 MT of sorghum per year. Tables 1 and 2 below indicate the amounts of sorghum and maize produced, respectively. These statistics, while from 2008, are unfortunately the most recent statistics available from the Ministry of Agriculture of Cameroon. Cameroon is a net exporter of sorghum, shipping its surpluses to neighboring countries like Chad and Nigeria. The total tonnage exported is not well captured by the authorities since they do not monitor transactions carried out by agricultural traders and much of the trade occurs outside of the formal marketplace. The amount of unofficial agricultural trade, although undocumented by the government, has been heavily researched by several in-country agricultural scholars who in 2008 estimated that 0.4 percent of annual GDP, amounting to 155,000 MT of agricultural goods, were exported via the unofficial, illegal marketplace. For more information on cross-border trade, please see the government interference section below. Maize produced in Cameroon is locally consumed by households and breweries as well as being used as animal feed. Some maize produced in Cameroon is exported to neighboring countries during the harvest season, including Chad, Nigeria, Gabon, the Republic of Congo, and Equatorial Guinea, but despite minor quantities of maize exports Cameroon remains a net importer of maize. Maize merchants from Chad and Nigeria, as well as processing companies such as Maiscam, the sole maize meal processor in Cameroon, procure cereals from the market. Demand for maize stocks is highest during the April to September lean season. Around three quarters of the beans and other pulses are produced in the



West and Northwest regions, with the harvest period occurring in July and August. For a graphic representation of Cameroon’s harvest calendar, see Figure 1. The specific type of beans WFP procured for this project was not mentioned in their final project evaluation report. With regard to cereal price fluctuations, the price of sorghum has been uncertain since 2010 due to the drought experienced in the north (for sorghum price variance over the project life cycle, see Table 3). Maize prices were also affected by the drought, and as a result have remained at high levels since 2011. Neighboring Sahelian countries of Chad and Nigeria, whose merchants frequently traverse the border to procure cereal commodities, were also affected, further driving up regional food insecurity and demand for any remaining stocks. The highest prices for maize are found during the April to September lean season (for price variance of maize see Table 4). Prices for rice, the second largest staple in the country, have remained stable. Rice is imported to satisfy demand in the southern regions, due to high transportation costs for moving cereals from producing regions in the north. In November 2010, the price of beans increased to its highest level as recorded by WFP project staff, due largely to merchants from the Republic of Congo, Equatorial Guinea, and Gabon purchasing prior to the Christmas season in December(for more on bean price variance, see Table 6).


The WFP project in Cameroon used hard tendering to procure the commodities. Ten suppliers submitted bids for sorghum, nine suppliers submitted bids for maize, and eight for beans.


The evaluation team’s analyses of available data for these three commodities shows no conclusive evidence of a market impact on the price of beans, maize, and sorghum. Please see Annex 3 and the market impact chapter of the main body of the report for further detail. Price data around the time of the LRP procurements are shown in Tables 3, 4, and 6 below. WFP reported that procurement for all eligible commodities took place from November to December 2010, during the start of harvest season, when prices normally fall to the lowest level. Due to a serious drought hitting the Far North region of Cameroon in 2009/2010, the production of sorghum for 2010 was lower than average and therefore the price of sorghum during the procurement stage remained high. The impact of WFP’s procurement, however, was not significant as it did not further increase the price of sorghum in this region. Based on the fact that maize prices remained stable during the procurement period, due to abundant availability, WFP’s procurement did not affect the market price in the North region. The most recent statistics available on regional production of sorghum and maize are from 2007 and 2008. Therefore a direct assessment of throughput for 2009 and 2010 cannot be made. However, if the 2007 and 2008 levels are averaged and used as a proxy, WFP’s three procurements would come to 0.05 percent of sorghum throughput, one percent of maize throughput, and 0.06 percent of beans throughput respectively. WFP procurements under the project are thus fairly small with relatively contained impact. WFP further reported that the price of beans in Bafoussam, the center city



of West region, increased significantly in November 2010, when the two suppliers initiated the procurement, after having signed contracts with WFP. The cause of this increase was attributed to strong demand from surrounding countries such as Gabon, the Republic of Congo, and Equatorial Guinea before the Christmas season, but WFP did not exclude the possibility that its procurement might have affected the market price further. The perceived impact of WFP’s procurement on the beans market was, nevertheless, considered to be limited given that the price of beans decreased the following month when the suppliers were completing the purchases in the local markets.


While Cameroon is well-suited to agricultural production in terms of terrain and climate, its import and export markets are skewed due to the agricultural policies of the government and a general lack of record-keeping for agricultural commodities passing through the border. A very significant amount of informal cross-border trade in agricultural products occurs, especially flowing toward Equatorial Guinea, Chad, and the Republic of Congo, which one researcher estimated at 0.4 percent of annual GDP in 2008, amounting to 155,000 MT of agricultural commodities.10 Prior to the country’s recession in the mid-1980s the government played a strong interventionist role in agricultural markets and in the establishment of subsidies and tariffs. It has since taken a more withdrawn approach following the onset of structural adjustment programs directed by the IMF and the World Bank. The National Cereal Board of Cameroon released its stocks of commodities when the prices rose significantly during the post-harvest season, thereby contributing to price stabilization. Since WFP’s procurement took place during the harvest season, when the price remained at its lowest, no interference occurred from the supply side. Import and export policies continue to pose problems to domestic farmers, however, with the trend over the last 15 years resulting in a massive upswing in food imports, particularly of rice, wheat, seafood and milk products. Massive imports of cereals particularly has led to severe consequences on consumption trends, with 20 percent of the national diet now supported by imported rice and wheat flour, in turn depressing economic livelihoods for local farmers whose goods are no longer as much in demand. Agricultural development aid to Cameroon has been declining since the 1990s, with a greater emphasis on increased local capacity for production through input and irrigation supplies, increased research and development, post-harvest protection, pest control, and access to agricultural financial services. A greater emphasis on strategic planning, monitoring, and evaluation is also illustrated in donor policy.11 Increasing public sector attention has turned to microcredit financing for small farmers in recent years, through the Credit Agricole Bank and the Financial and Management Accountancy Centre (FIMAC) project.

10 http://www.theigc.org/sites/default/files/presentation_slides/nkendah.pdf 30. 11 http://www.oecd.org/dataoecd/38/58/41041414.pdf p. 20

http://www.theigc.org/sites/default/files/presentation_slides/nkendah.pdf



Unfortunately both of these measures have failed to produce meaningful results due to endemic corruption in the national government.12 In 2010 the U.S. government supplied Cameroon with 3,350 MT of rice under a Title II emergency program and 1,130 MT of vegetable oil, rice, and beans under a McGovern-Dole International Food for Education and Child Nutrition program.13 Another 3,000 MT was programmed to be shipped through USAID and 1,100 MT was programmed to be shipped through the McGovern-Dole Food for Education and Child Nutrition program in 2011.14


WFP Cameroon procured 329 MT of sorghum, 1,790 MT of maize grain, and 543 MT of beans under the project, all via hard tender.


The average overall procurement time for three procurements was 109 days; 16 days elapsed in contracting, and 93 for delivery. The average overall time for sorghum was 133 days, with 15 for contracting and 118 for delivery. The average overall time for maize was 110 days, with 15 for contracting and 95 for delivery. The average overall time for beans was 84 days, with 19 for contracting and 65 for delivery.

VI Total cost of procurement, including storage, handling, transportation and administrative costs

Average commodity costs per MT were as follows: • Sorghum: $334.54/MT; • Maize: $408.48/MT; and • Beans: $1,328.54/MT.

TSH costs for all three procurements amounted to $309.05/MT. ii. Assess: I Whether the requirements of this

section have been met WFP reporting was found to be compliant with requirements. Sections of the final report regarding impact on markets, producers and beneficiaries were superficial but addressed.

II Impact of different methodologies and approaches on: aa Local and regional agricultural

producers, including large and small agricultural producers

WFP reported having bought the commodities from traders who collected the commodities from small-scale farmers in the local market. The local purchase had a positive impact on these farmers, increasing their income particularly as this purchase took place immediately after the harvest. The evaluation team cannot comment as to the relative benefit to the traders vs. small farmers, but undoubtedly the procurement provided a source of income and increased demand for these small farmers as well as the traders aggregating such surplus harvests.

bb Markets After reviewing conflicting evidence from the final report and examining possible price changes through analysis of short-term elasticities, the evaluation team recognizes the possibility that the LRP project in Cameroon may have caused a market impact, but cannot confidently attribute a rise in prices to the LRP project. However, this basic assessment of elasticities contradicts the WFP final project report. WFP’s report indicated no impact on sorghum or maize prices, but was unable to rule out impact on prices for beans. An increase in bean prices was attributed mainly to high demand from surrounding countries during the Christmas holiday season, but this demand diminished after one month.

12 http://www.rtfn-watch.org/uploads/media/ANoRF_Cameroon_-_Country_Report_on_the_Right_to_Food.pdf p. 24. 13 United States Agency for International Development and United States Department of Agriculture. U.S. International Food Assistance Report 2010. N.d. Web. 14 Apr. 2010. <http://www.usaid.gov/our_work/humanitarian_assistance/ffp/fy2010.ifarreport.pdf>. 14 US Food Aid Tables, FY 2011, < http://www.fas.usda.gov/excredits/FoodAid/Reports/reports.html >.



Please see Annex 3 and the market impact chapter of the main body of the report for further detail. WFP also reported a positive impact on secondary actors in the markets such as bag manufacturers, fumigation and quality testing companies, and transporters. Local procurement via hard tender also enhanced competition in the market as well as familiarized large and small agricultural producers with standard contracting processes. With regard to the levels of market integration, Cameroon provides significant levels of agricultural commodities to neighboring countries, as described previously. Market integration within the country itself is decent but reflects infrastructure difficulties in transporting goods over long distances. Graphs of retail prices for maize and wheat over the January 2005 to December 2011 period are detailed in Figures 2 and 3 and illustrate the capital versus regional market divide. Wheat was not procured as part of the project, but helps to illustrate varying levels of regional market integration for commodity types. Douala is the primary external port city, sitting on the Gulf of Guinea, and thus presents a better reflection of external and international price data. Garoua is located in the north on Bénoué River, and thus prices reported from this location reflect not only increased prices due to drought in the North, but also increased demand from neighboring Nigeria and Chad. The capital, Yaoundé, is in the coastal plains region and thus provides strong indication of in-country prices. Maize prices are notably higher in the port city of Douala versus in-country in Garoua, suggesting that demand in the northern regions is lower than in Douala, likely because of regional demand for goods supplied via oceanic freight. Surpluses and deficits due to seasonal rainfall shortages correspond between the two cities’ data, suggesting that the regional demand is more of a factor in the higher price of maize in Douala as opposed to weather-related supply issues. The higher retail price of wheat in the south-central region versus in the port city of Douala again reflects regional commodity flows. The South and East regions of the country are more heavily dependent on cereal imports like wheat for nutrition.

cc Low-income consumers WFP reported that the impact on low-income consumers is considered to be very limited because the prices of the commodities remained stable and the availability was abundant during WFP’s procurement period. As described earlier, when using a proxy to determine 2009-2010 throughput, WFP procurements only amounted to 1 percent and 0.06 percent of the regional output for maize and beans, respectively, making it difficult to attribute a rise in prices for these two commodities to LRP. However, as price changes were witnessed, the impact of the LRP project cannot be ruled out entirely. As market price impact for price changes in maize and beans cannot be confidently attributed to the LRP project, any impact on low-income consumers cannot be rightfully attributed either.

dd Program recipients The recipients of the project were drought-affected people in the Logone and Chari regions in the Far North of Cameroon. Two hundred sixteen thousand targeted beneficiaries received food through general food



distributions. Given the large distance between the recipients’ homes and the purchase market as well as the limited purchasing power of the recipients, the distribution of commodities was a critical reaction to food insecurity.

III Time of delivery For the three procurements, there was an average overall time for procurement of 109 days with 16 days for contracting and 93 days for delivery. The breakdown of procurement times by commodity is as follows: Sorghum procurements took a total of 133 days, with 15 for contracting and 118 for delivery; maize took a total of 110 days, with 15 for contracting and 95 for delivery; and beans took a total of 84 days, with 19 days spent in the contracting phase and 65 in delivery.

iii. Compare: Quality and safety assurances Atlantic Survey Company was contracted to provide quality and quantity

testing and inspection of the commodities by performing visual and humidity testing such as looking for broken and discolored grains, insect damage, the presence of foreign materials, odors, etc. All goods were found to be in order, with very minor defects as described above at acceptably low levels (typically less than 1%). Atlantic Survey Company was also contracted to fumigate the commodities.

TABLE 1: SORGHUM PRODUCTION IN CAMEROON (MT)

Regions 2007 2008 Far North 622,540 621,253

Others 353,682 385,225

Total 976,222 1,006,478 Source: Ministry of Agriculture of Cameroon database, in USDA Local and Regional Procurement Pilot Project, Evaluation Report –Final LRP Project Report World Food Programme/Cameroon

TABLE 2: MAIZE PRODUCTION IN CAMEROON (MT)

Regions 2007 2008 Far North 133,706 159,752

North 172,538 173,938

Others 1,015,916 1,061,142

Total 1,322,160 1,394,832 Source: Ministry of Agriculture of Cameroon database, in USDA Local and Regional Procurement Pilot Project, Evaluation Report –Final LRP Project Report World Food Programme/Cameroon


TABLE 3: RETAIL PRICE MOVEMENT OF SORGHUM IN THE FAR NORTH REGION

(PER 100 KG)

Year 2011 2010 2009

Market Month/Currency FCA USD FCA USD FCA USD

Far North Region/ Maroua

January 13,000 27.00 12,000 25.90 12,000 23.70

February 13,000 27.10 11,000 22.80 13,000 25.40

March 13,000 28.10 12,000 24.70 12,500 25.90

April 14,000 31.00 12,000 24.30 13,000 25.80

May 14,000 30.20 12,000 23.00 13,000 27.80

June 12,500 23.50 14,000 29.50

July 13,000 25.60 14,000 30.40

August 13,500 26.20 13,000 28.40

September 14,000 28.60 11,000 24.80

October 14,000 29.7 8,000 18.30

November 14,000 29.0 8,000 18.10

December 14,000 27.9 8,000 17.40 Calculations based on prevailing currency exchange rates

Source: Ministry of Agriculture of Cameroon database, in USDA Local and Regional Procurement Pilot Project, Evaluation Report –Final LRP Project Report World Food Programme/Cameroon

TABLE 4: RETAIL PRICE MOVEMENT OF MAIZE IN NORTH REGION (PER 100 KG)

Year 2011 2010

Market Month/Currency FCA USD FCA USD

North Region/ Garoua

January 11,500 23.90 13,000 28.10

February 12,500 26.10 13,500 28.00

March 13,500 29.20 14,000 28.80

April 14,000 31.00 14,000 28.40

May 17,000 36.70 14,000 26.80

June 16,000 34.60 13,000 24.40

July 14,750 29.00

August 14,500 28.10

September 13,000 26.60

October 11,000 23.30

November 11,000 22.80

December 11,000 21.90 Calculations based on prevailing currency exchange rates



TABLE 5: BEAN PRODUCTION IN CAMEROON (MT)

Regions 2007 2008 West 81,285 87,015

North-West 93,307 91,638

Others 84,819 91,989

Total 259,411 270,642 Source: Ministry of Agriculture of Cameroon database, in USDA Local and Regional Procurement Pilot Project, Evaluation Report –Final LRP Project Report World Food Programme/Cameroon

TABLE 6: RETAIL PRICE MOVEMENT OF BEANS IN WEST REGION (PER 100 KG)

Year 2011 2010

Market Month/Currency FCA USD FCA USD

West/ Bafoussam

January 54,620 113.40

February 43,000 89.60

March 52,000 112.40

April 52,850 117.00

May 56,900 122.70

June 54,520 117.80

July 43,500 85.70

August 43,600 84.60

September 45,900 93.80

October 43,300 91.90

November 58,400 121.00

December 40,000 79.70 Calculations based on prevailing currency exchange rates



FIGURE 1: CAMEROON HARVEST CALENDAR

FIGURE 2: MAIZE RETAIL PRICES, PER CFA FRANC

Source: FAO GIEWS Price Tool at http://www.fao.org/giews/pricetool2/


FIGURE 3: WHEAT RETAIL PRICES, PER CFA FRANC

Source: FAO GIEWS Price Tool http://www.fao.org/giews/pricetool2/


CHAD

WFP REGIONAL PROCUREMENT PROJECT, EMERGENCY

Through the LRP grant of $3.1 million, WFP purchased some 2,600 MT of maize and pulses, in Cameroon and Niger, respectively, for distribution in Chad. These regional purchases were made through hard tenders and were distributed to beneficiaries in affected regions of the Sahelian belt of Chad under Emergency Operation (EMOP) 200112. Specifically, the project procured 1,111 MT of cowpeas in Niger; and 1,512 MT of maize in Cameroon. The project distributed food to 737,000 individuals.

Reporting Requirement Notes i. For each market & commodity procured, describe:


Agricultural markets in Chad are divisible into two different zones of east and west which are more integrated with their neighboring countries than they are with each other. The western basin has fairly integrated markets internally, as well as with the neighboring countries of Cameroon and Nigeria. The eastern zone is much less integrated with the rest of the country and is influenced by market prices in neighboring Sudan.15 WFP reported that in January 2011, a few months before WFP made the USDA-funded regional purchases in Cameroon and Niger, prices of cereals indicated a downward trend in Chad and Niger, both of which recorded significant production increases for the 2010 harvest, in what was termed a bumper year.16 Moreover during the same period, and earlier in 2010, the price of cereals reached record highs in the Sudanese province of Darfur, which neighbors eastern Chad. Transaction cost analysis done by Famine Early Warning Systems Network (FEWSNET) indicated that even during the lean period in eastern Chad, high prices in Darfur created strong incentives to export cereals across the border to Sudan, which only exacerbated pressures on prices and food insecurity in eastern Sahelian Chad. Finally, the spread between millet prices in the western zone and eastern zone doubled in 2010, indicating that not only were cereal prices erratic during this time period, but illustrating that market integration between the two Chadian market zones was deteriorating and that any western region surpluses may not make it to deficit regions in the east. With regards to the source countries for these regional purchases, cowpeas were sourced in Niger and maize in Cameroon. Niger grows cowpeas for commercial export. A 2010 survey by the Ministry of Agriculture in Niger confirmed that cash crops such as cowpeas had exceptional harvests. Production of cowpeas registered a significant increase with production estimated at 1,972,826 MT in 2010, against 1,543,943 MT in 2008 and 787,472 MT in 2009, making it an excellent option for regional pulse procurement for neighboring deficit countries such as eastern Chad.

15 World Food Programme. “Executive Brief: Markets and Food Security in Chad.” February 2011. Web. 6 May 2012. <http://home.wfp.org/stellent/groups/public/documents/ena/wfp231403.pdf> 16 USDA Local and Regional Procurement Pilot Project, Evaluation Report Emergency Food Assistance to Drought-Affected Populations in Chad: WFP Chad Final LRP Project Report, questions/further clarification items. Unpublished. Chad: WFP Chad Country Office, 2011. 6.


Reporting Requirement Notes Maize production in Cameroon in 2007 and 2008 is shown in Table 1. No information was found for historical market demand, but historical prices of maize in Chad are presented in Figure 1. WFP reported that historical data from 2006 to 2011 show significant price fluctuations for beans in Niger. High prices were observed in 2010, especially between June and August. The price increase was almost 50 percent compared to bean prices in 2006 and 2007 during the same period.


WFP reported that in Niger, eleven suppliers were invited to bid to supply 1,111 MT of beans. WFP received ten responses out of which two suppliers were selected based on their prices being the most competitive. In Cameroon, WFP procured 1,512 MT of unprocessed maize. Seven offers were received via a competitive tender. Upon review of the offers, a contract was awarded to two suppliers.


WFP reported that although no official statistics on maize production in Cameroon for 2010 are available, information from the WFP Cameroon Country Office indicates that prices remained stable during the procurement period. This country office collects and processes information from local farmers and suppliers. It therefore appears that WFP’s procurement did not affect the market price in the North region. WFP further reported that the price of beans in Niger showed a 3 percent increase between the week before the purchase and the day of purchase, while a 6.6 percent increase was recorded between the day of procurement and one week after procurement. WFP reported that in 2010 Niger produced 1,773,423 MT of beans and it was estimated that 300,000 MT of this crop were exported. Therefore, the purchase of 1,111 MT (0.37 percent of the market quantity) of beans for Chad would have had an insignificant impact on the market. WFP also reported the wholesale and retail (producer) prices of cereals in Chad were fixed in 2008 by the Government Ministry of Commerce and Industry as a response to rising prices, according to Governmental directive No. 030/PR/PM/MC 2010, yet in early 2011 when the WFP procurements took place these price ceilings still existed, creating an imbalance of supply and demand and depressing traders’ interest in trading cereals and producers’ interest in increasing supply. Consumer prices for cereals in Chad remained the same one-week before procurement, on the date of procurement, and one week after procurement.


The evaluation team found that since the food price increases of 2008, the host government has implemented price controls and export restrictions. These price controls – put into place in 2008, in 2009, and at the onset of the food crisis of 2010 – have limited traders’ interest for trading in grains. WFP reported that the Chadian government set up a food security stock in response to the price increase of 2008. This stock was managed by the Office National de la Sécurité Alimentaire (ONASA), the agency also responsible for implementing price controls and export restrictions. During the specific time of the LRP project, procurement price controls were still in place and regulated by the Government. The prices of cereals fixed by the government varied between


Reporting Requirement Notes 220.8 Central Africa Francs (XAF)/kg ($0.42USD/kg) and 229 XAF/kg ($0.43 USD/kg,17 depending on the production quantity and the accessibility of various producing regions. 4,800 MT of bulgur wheat and bread flour were sent to Chad through Title II funding in 2010, along with 1,130 MT of vegetable oil and CSB through the McGovern-Dole International Food for Education and Child Nutrition Program.18 An additional 61,400 MT of commodities were programmed to be shipped to Chad by USAID for distribution in 2011.19


The project used a hard tender to procure 1,111 MT of cowpeas and 1,512 MT of maize.


The average procurement time for maize was 106 days overall, with 40 days for contracting and 66 days for delivery. The average procurement time for cowpeas was 51 days, with six days for contracting and 45 days for delivery.


Average commodity costs per MT prices were as follows: • Maize cost $356.50 per MT, with an additional $455.16 per MT required

for TSH. • Beans cost $566.30 per MT, with an additional $455.16 per MT for

TSH.


this section have been met WFP reporting was found to be compliant with requirements, however WFP reported on cereals (type unspecified) for Niger while maize was procured from Cameroon. Data on maize production in Cameroon are presented only for 2007 and 2008. Historical data on market price fluctuations are provided for maize in Chad (2002 to 2010) and beans in Niger (2006 to 2011) below. No similar information was found for beans in Chad or maize in Cameroon. No data were found for historical market demand. Minimal information is available regarding the impact on producers, markets, low-income consumers, and beneficiaries.



WFP reported that regional purchases had a positive impact on farmers (local and regional agricultural producers) by increasing their income. WFP further reported that although no extensive assessments were done, LRP purchases had a positive indirect impact on improving food security by encouraging farmers to increase production. WFP’s report goes on to say that hard tenders via large traders who aggregate the surpluses of small scale farmers may allow these small farmers to sell their produce at the right time and right price20, permitting them to prepare for the next farming season with enough income to purchase the necessary seed stock, fertilizer and other equipment. The evaluation team is unable to comment

17 FEWSNET. Mali Monthly price bulletins for February 2010 and February 2011. 18 United States Agency for International Development and United States Department of Agriculture. U.S. International Food Assistance Report 2010. N.d. Web. 14 Apr. 2010. <http://www.usaid.gov/our_work/humanitarian_assistance/ffp/fy2010.ifarreport.pdf>. 19 US Food Aid Tables FY 2011 < http://www.fas.usda.gov/excredits/FoodAid/Reports/reports.html>. 20 USDA Local and Regional Procurement Pilot Project, Evaluation Report Emergency Food Assistance to Drought-Affected Populations in Chad: WFP Chad Final LRP Project Report. Unpublished. Chad: WFP Chad Country Office, 2011. 6.


Reporting Requirement Notes on the relative benefit between the large traders versus small farmers, only to say that both played an important role in the supply chain of these commodities to those in need. Additionally, procuring regionally allows for improvement of market integration, even on a limited scale.

bb Markets WFP reported that its procurement methodology had a positive impact on the markets since it involved other actors including bag manufacturers and companies dealing with fumigation of food and transport vehicles. The methodology is also seen by WFP to enhance competition in the market. The project procured 1,111 MT of cowpeas in Niger; and 1,512 MT of maize in Cameroon. The evaluation team’s assessment of market impact indicated that an impact from the regional procurement of maize and cowpeas for the project was very unlikely since cowpea prices fell in procurement zones in Niger. For maize, prices rose uniformly over the procurement period, with the increase starting prior to procurement and continuing after it, insinuating that regional procurement had no impact and that maize price rises were of a seasonal nature. Please see Annex 3 and the market impact chapter of the main body of the report for further detail.

cc Low-income consumers

WFP reported the impact on low-income consumers as very limited in both Cameroon and Niger. This is likely due to stable prices and abundant availability during the timeframe for procurement.

dd Program recipients The 586,745 project beneficiaries were located in the drought-affected Sahelian western belt of Chad. The commodities are distributed in the regions of Mao, Mongo, Ati, Bol, Mossoro and Hadjer Lamis through targeted general food distributions to families of the beneficiaries of the blanket feeding rations to 6-23 months children and lactating women, in order to protect (and avoid intra-household sharing of) those supplementary individual rations to the most at-risk population groups during the lean season as to prevent a rise in acute malnutrition in the regions of the Sahelian belt.

III Time of delivery The average procurement time for maize was 106 days overall, with 40 days for contracting and 66 days for delivery. The average procurement time for cowpeas was 51 days, with six days for contracting and 45 days for delivery.

iii. Compare: IV Quality and safety assurances The evaluation team reviewed examples of Storage Site Inspection Reports

conducted by the Laboratoire National de Santé Publique et d’Expertise (LANSPEX) and by Africa Marine Surveys (AMS) Cameroon, and a sample document of laboratory testing from AMS. All documents reviewed suggested no major issues with food quality or safety.


FIGURE 1: RETAIL MAIZE PRICES IN N’DJAMÉNA (CAPITAL OF CHAD), 2002 – 2010

Source: FEWSNET

FIGURE 2: TRENDS OF BEAN PRICES IN NIGER FROM 2006 TO 2011

Source: WFP Chad Final LRP Project Report Questions/Further Clarification Items

TABLE 1: MAIZE PRODUCTION IN CAMEROON (MT)

Region Maize

Year 2007 Year 2008

Far North 133,706 159,752

North 172,538 173,938

Others 1,015,916 1,061,142

Total 1,322,160 1,394,832

Source: Ministry of Agriculture of Cameroon Database, in USDA Local and Regional Procurement Pilot Project, Evaluation Report Emergency Food Assistance to Drought-Affected Populations in Chad: WFP Chad Final LRP Project Report

0

200

400

600

800

CFA

fran

cs p

er k

ilogr

am

-

100,000

200,000

300,000

400,000

500,000

Jan

Feb

Mar Ap

r

May Jun Jul

Aug

Sept Oct

Nov De

c

Price (XOF)

2006 2007 2008 2009 2010 2011


REPUBLIC OF CONGO

WFP REGIONAL PROCUREMENT PROJECT, EMERGENCY

WFP implemented this project in the Republic of Congo in response to an emergency that occurred following inter-ethnic clashes in Equateur Province in the Democratic Republic of Congo (DRC). This emergency displaced many people, sending them fleeing across the Ubangui River to the Likouala department of the Republic of Congo. The project’s two goals were to: i) improve food consumption and security; and ii) eliminate acute malnutrition among children under five.

WFP purchased 1,458 MT of rice in the Bumba region in Equateur Province of the DRC. Rice was procured through a hard tender; winners were two local commercial traders. Each trader ma de deliveries across the river to two WFP warehouses from which the rice was dispatched to smaller warehouses along the Ubangui River on the Republic of Congo side for direct distribution to 114,000 refugees. The tender was issued on April 5, 2011, and the last distributions occurred in October 2011.



Prior to the civil war in the late 1990s, several large commercial traders in the DRC processed and marketed more than 500,000 MT of rice per year. However, during and following the civil war, deterioration in economic and social infrastructure in the procurement region reduced the land area under cultivation and caused a reemergence of subsistence farming in several areas. Because little reinvestment in commercial agriculture took place after the civil war, annual rice production now ranges between 80,000 to 95,000 MT. It is important to note that food deficit regions of the Republic of Congo, including those receiving the commodities procured through the project, and the food surplus of the DRC regions supplying the project, were much more integrated in the marketplace prior to the civil war. Demand for rice in the DRC is not very high as it is grown mostly as a cash crop, and not consumed regularly by subsistence farmers. Rice is considered a food only consumed on special occasions and for ceremonies, thus local demand is rather thin. WFP’s final project report indicated that local demand is estimated at 24,300 MT per year, with peaks in demand in July. As a consequence, price movements show seasonality with a peak in July and August, corresponding to the time just before the harvest. After the August harvest, rice prices drop from September to November and then increase again. The retail price varies considerably depending on the market: in Bumba, in September 2011, prices ranging from 400 to 600 Congelese Franc/Kg were recorded.


The WFP Country Office in the DRC launched a hard tender, which they refer to as a selective tender, on behalf of the WFP Country Office in the Republic of Congo on April 5, 2011, for the supply of 1,500 MT of rice. The two main food suppliers in Equateur Province sent bids. The procurement process respected the WFP Food Procurement Manual Section Two (a set of procurement rules used by WFP offices), and related standards of conduct.

II Impact of the procurement of the eligible commodity on producer and consumer

The evaluation team’s analyses of existing and available data indicate that this project had no likely impact on the prices of the affected commodities. Please see Annex 3 and the market impact chapter of the main body of the report for further detail.


Reporting Requirement Notes prices in the market

WFP in the Republic of Congo estimated that it was difficult to assess the impact of its procurements on the producers’ welfare, since 1,480 MT was a small amount for the area, amounting to only 1.5 percent of the average annual production level of 95,000 MT per year (see the above section on prevailing supply). In addition, WFP explained that farm-gate prices vary between 150 and 200 Congolese Francs/Kg; that farmers are often paid in-kind, and that different remuneration modalities are used by large traders. Farmers interviewed by WFP explained that the farm-gate price increased by 30 Congolese Francs/Kg on average. WFP reported that prices are fixed by commercial traders at the beginning of the campaign between such traders and local suppliers and small farmers in order to establish the purchase price before harvest quantities are known. These traders’ use of anticipatory market prices and not actual market prices at harvest provide a means for producer commodities prices to remain rather stable over the harvest period and allow farmers and traders to set expectations at the outset of a harvest season. As producer price fixing is standard practice between large traders and small suppliers in the area, the WFP procurement could not have had any impact on those prices. WFP in the Republic of Congo explained that rice was not widely consumed in the procurement area, with only one quarter of the rice produced in the region consumed locally and a fraction of it sold on the local markets. As such, WFP staff expected that rice would show high price elasticity for consumers in the procurement zone, expecting that prices would rise significantly in response to WFP’s rice procurements. However, as local demand was fairly low, WFP also anticipated that many consumers would ‘remove’ themselves from the marketplace in response to rising rice prices. In addition, the procurement amounts represented below 2 percent of annual production as explained earlier, suggesting that any effect on price would be small and short-lived. WFP reported that due to high price elasticity, a drastic drop in rice prices could be expected in distribution zones with any injection of supply. However, WFP indicated that this expected fall in price was not actually experienced and instead rice prices rose slightly. WFP staff noted that since much of the rice was consumed in the distribution zones, and thus prevented from re-entering the market place, the anticipated price decrease effect was buffered by consumption. No seasonal price fluctuation was witnessed in distribution markets. Such fluctuation was not expected since most rice is brought in from outside the region and the country, and is not affected by local harvest supply shocks on the whole. Thus, WFP and the evaluation team concur that no significant impact was evident on either consumer or producer prices due to WFP’s regional rice procurement.


WFP did not report any formal government intervention influencing farm-gate prices, but did note that farmers reported petty corruption by officials and roadblocks by the police. Such costs are increases in farmers’ transaction costs from interference at the hands of state bureaucrats and border guards rather than through regulation. Title II emergency activities included the shipment of 2,980 MT of rice, vegetable oil, and yellow split peas to the Republic of Congo in 2010.21

21 "U.S. International Food Assistance Report 2010." USAID Food For Peace. USAID, n.d. Web. 9 Apr. 2012. <http://www.usaid.gov/our_work/humanitarian_assistance/ffp/annrep.html>.


Reporting Requirement Notes IV Quantities and types of

eligible commodities procured in the market

WFP procured 1,458 MT of rice from two local commercial traders through a hard tender, or as WFP refers to it, a selective tender.


Two rice procurements were obtained via the same tender, but awarded to two different vendors. Each took 112 days total with 55 days for contracting and 57 days for delivery.


Two contracts for rice yielded different prices, as follows: The first procurement of 958 MT cost $839.60 per MT for the commodities and the second procurement of 500 MT cost $687.50 per MT. TSH costs for both procurements averaged $403.77 per MT.

ii. Assess: I Whether the

requirements of this section have been met

WFP reporting was found to be compliant with requirements, however no information was found regarding historic market demand or price movements.



As mentioned earlier, large traders in the DRC procured goods through small farmers by fixing producer farm gate prices prior to the harvest period, providing an opportunity for both parties to hedge their bets on the likely outcome of the harvest and thus future prices. Price fixing up the supply chain for these rice procurements does not reflect on WFP’s procurement, but rather on local market conditions and practices. WFP also reported that small farmers were constrained by their physical access to the market and to marketing tools, including but not limited to, the inadequate and irregular flow of supplies such as bags, transportation, spare parts and fuel, as well as the lack of basic institutions and government-furnished common goods such as access to water, electricity and passable roads.

bb Market

The evaluation team’s analyses of existing data indicate that this project had no likely impact on the prices of the affected commodities in the procurement areas. Please see Annex 3 and the market impact chapter of the main body of the report for further detail. WFP noted a minimal increase in retail rice prices in one of the two distribution areas (Impfondo) over the months of August and September, the distribution months, although the increase was probably less than during the usual peak typical of the season. In the other distribution area (Bétou), retail rice prices were constant overall during the same months.22 Price changes in Impfondo a week after distribution showed increases for two distributions and decreases for the remaining two; however, this is not inconsistent with an overall increase observed over two months during the distribution period as it only shows short-term changes. Price changes the week after distribution were not available for the Bétou distribution area. This price drop lasted for a few days and countered the impact of the normal seasonal increase. Since rice is not the staple product in the area, the volumes traded are small. Focus groups with vendors found that they sensed no difference in terms of quantities sold when compared to previous years. Prices in the distribution region follow a seasonal pattern that is different from Bumba since the two markets are not integrated. Rice is available all year long, but the price

22 See figure 5 below for a graphical representation of monthly prices in the two distribution areas.


Reporting Requirement Notes peaks (February/March and August) are due to heavy rains and the ensuing poor road conditions caused by high water levels on the river.

cc Low-income consumers The minimal price fluctuations discussed in the “markets assessment” above, and reported by WFP, may have affected retail prices for low-income consumers, but as previously mentioned, the minor price rise in distribution zones appeared to have counteracted the normal seasonal increase. The evaluation team’s analyses detected no likely impact on prices.

dd Program recipients Some 117,000 refugee beneficiaries were provided with rice and thereby assisted in their survival. Once distributions started, the evaluation team learned from first-hand accounts and interviews conducted during the project site visit that refugees began selling a portion of their rice in exchange for cash to buy other commodities sold on the markets such as dried fish and oil. They in turn would sell some of these goods on other markets in the area. The ratio of rice sold by refugees in comparison to rice already present in the area was impossible to determine because these transactions were not recorded. Note that the reselling of food aid commodities, for aid that has been sourced locally regionally or internationally, is not an indication that commodities were not needed. Several explanations can be attributed to this action on the part of the beneficiary, often linking back to a desire by the participant for greater variety in the diet.

III Time of delivery Two rice procurements originating from the same tender, but awarded to two vendors took 112 days total to complete, with 55 days needed for contracting and 57 days for delivery.

iii. Compare: IV Quality and safety

assurances One of the selection criteria for the suppliers was their ability to meet quality standards for rice. The participant provided documentation on testing for moisture, losses, and insects found in the commodities. Examples of laboratory test results were reviewed by the evaluation team and no food quality or safety issues were found. Testing was performed by the Democratic Republic of Congo’s Control Office and Department of Laboratories for both vendors.


FIGURE 1: RETAIL PRICES ON THE BUMBA MARKET FOR RICE

Source: Fossi, Filipo. USDA Local and Regional Procurement Pilot Project, Evaluation Report–Final LRP Project Report World Food Programme/Republic of Congo. Unpublished. Brazzaville, Republic of Congo: WFP/RoC Country Office, 2011

TABLE 1: PRICES IN PROCUREMENT AREA AT SHIPMENT TIMES (CONGOLESE FRANCS)

Time Retail price in Bumba Variation from baseline

One week prior to shipment (baseline) 650

1st shipment: July 13th 650 0%

One week after shipment 630 -3%


2nd shipment: August 18th 570 -5%

One week after shipment 570 -5%


3rd shipment: Sept. 12th 570 0%

One week after shipment 450 -21% Source: Fossi, Filipo. USDA Local and Regional Procurement Pilot Project, Evaluation Report–Final LRP Project Report World Food Programme/Republic of Congo. Unpublished. Brazzaville, Republic of Congo: WFP/RoC Country Office, 2011

0

100

200

300

400

500

600

700

Jun-11 Jul-11 Aug-11 Sep-11

pric

e (C

ongo

lese

Fra

ncs)

Jun-11 Jul-11 Aug-11 Sep-11Bumba (average) 600 650 570 400


FIGURE 2: PRICE SERIES OF LOCAL AND IMPORTED RICE IN MBANDAKA AND KINSHASA23


FIGURE 3.A: MONTHLY PRICES OF LOCAL RICE IN MBANDAKA


23 Mbandaka and Kinshasa are located in the DRC and are the main destinations of rice produced in Bumba

0

200

400

600

800

1,000

1,200

1,400

1,600

1,800

pric

e (C

ongo

lese

Fra

ncs)

Imported Rice

Mbandaka

Kinshasa

0

200

400

600

800

1,000

1,200

1,400

Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov

Reta

il pr

ice

(Con

gole

se F

ranc

s)

2009

2010

2011


FIGURE 3.B: MONTHLY PRICES OF LOCAL RICE IN MBANDAKA AND KINSHASA


0

200

400

600

800

1,000

1,200

1,400

reta

il pr

ice

(Con

gole

se F

ranc

s)

Local Rice

Mbandaka

Kinshasa

Mbandaka(smoothed)


FIGURE 4: FORECASTED AND OBSERVED PRICES

Month

pric

e

agoaprdicagoaprdicagoaprdic

1750

1500

1250

1000

750

500

Variable

LAforeUAfore

local riceAforecast

ARIMA Model of Price of local rice in Mbandaka

Source: Fossi, Filipo. USDA Local and Regional Procurement Pilot Project, Evaluation Report–Final LRP Project Report World Food Programme/Republic of Congo. Unpublished. Brazzaville, Republic of Congo: WFP/RoC Country Office, 2011 Note: months listed on this figure correspond to December, April, and August.

FIGURE 5: RETAIL PRICES IN IMPFONDO AND BETOU (DISTRIBUTION AREAS)


0

100

200

300

400

500

600

700

Jan-11 Feb-11 Mar-11 Apr-11 May-11 Jun-11 Jul-11 Aug-11 Sep-11

CFA

Fran

cs

Impfondo

Betou

Price Congol-ese Francs


TABLE 2: PRICES IN DISTRIBUTION AREAS

Time Impfondo Variation from Baseline

One week prior to distribution (baseline)

700

Distribution 700 (Aug. 8) 0%

One week after distribution 650 -7%


650

Distribution 625 (August 12) 4%

One week after distribution 625 -4%


635

Distribution 650 (Aug. 20) +2.5%

One week after distribution 645 +1.6%


640

Distribution 645(Sept 19) +1%

One week after distribution 645 +1% Source: Fossi, Filipo. USDA Local and Regional Procurement Pilot Project, Evaluation Report–Final LRP Project Report World Food Programme/Republic of Congo. Unpublished. Brazzaville, Republic of Congo: WFP/RoC Country Office, 2011


GUATEMALA

CRS LOCAL PROCUREMENT PROJECT, EMERGENCY

CRS used funds from the USDA LRP Project to procure 1224.47 MT of white maize, 146.94 MT of beans, and 159.47 MT of corn and soy fortified flour, known in Guatemala as Incaparina. All commodities were procured and distributed in Guatemala. All three commodities were procured through hard tenders and were distributed directly to project beneficiaries. The project also included a Food-for-Work component for some of the beneficiaries. Farmers’ associations and private suppliers both supplied commodities to the project. The project distributed food to 3,000 families in the Department of Santa Rosa.



Historic maize production, seen in Table 1, shows an estimated 2010/2011 national production level of 1.7 million MT (35.8 million of quintals, of which 32.2 million are white maize and 3.6 million are yellow maize. Historic bean production, seen in Table 2, shows production of over 200,000 MT per year since the 2004/2005 season. Data for 2010 from FAOSTAT imputes a figure of 181,500 MT. No historic market supply data was found for fortified flour. Information on the historic consumption of maize and beans are provided in Tables 3 and 4 respectively. Maize consumption increased between the 2004/2005 and 2008/2009 agricultural cycles and decreased between the 2008/2009 and 2009/2010 cycles. They then rose again slightly in 2010/2011. CRS reported that apparent consumption of beans has increased every year at a rate of 1.6 percent per year. No historic market demand data was found for fortified flour. Retail maize prices in Guatemala typically follow an annual cycle with the lowest prices in harvest months and highest prices during the rainy season. Harvest months vary by region with two harvests in tropical zones, the first from August to September and the second from November to December. In the North and South zones the first harvest is in October and the second from February to March. Rainy seasons occur from May through mid-July and mid-August through the end of October. Prices increased in January-February 2011 and then again in May 2011 (Figure 1), consistent with seasonal patterns. Retail black bean prices show a similar seasonal pattern to that of maize, with prices following a relatively stable annual pattern until a significant increase early in 2011 (Figure 2). Tables 5, 6, and 7 provide further historic price data for maize, beans, and fortified flour respectively. CRS reported that prices of fortified flour (Incaparina) have not suffered abrupt changes during the past decade. Price increases shown in the past few years in Table 7 are due to the high cost and/or availability of ingredients (e.g. maize and soy) as well as increases in global gas prices.


CRS used a competitive tendering process to procure commodities. Potential farmers association and private suppliers were identified through WFP references and responses to newspaper ads. Farmers associations were given preference, but it was determined that many could not meet one or more of the three


Reporting Requirement Notes prerequisites for qualification. Organizational capacity, registration with the Guatemalan Superintendent of Tax Administration, and the ability to provide high quality goods, packaging and storage capacity were deemed to be the three most important qualifications for bidding organizations.


CRS reported that according to the food price analysis performed by Cornell University, CRS’ local purchases of beans, white maize, and Incaparina under the USDA LRP Project had no economically or statistically significant effect on retail food prices in Guatemala. Cornell University estimated the effects of the project purchases on the retail prices of yellow maize, a substitute for the white maize procured by the program. CRS’ local maize procurement actions were not associated with statistically significant changes in yellow maize prices. Similarly, CRS’ local maize procurements were not associated with statistically significant changes in the retail prices of white maize. Point estimates of the effects of CRS’ local procurement activities on retail bean and Incaparina food prices are typically ± 1 percent or less. CRS further reported that the same conclusions apply when focused exclusively on retail prices in markets within the procurement and distribution zone. In this zone LRP activity had no statistically significant effects on prices, indicating that these activities did not adversely affect normal market functioning in the geographic region within which the CRS activities took place. The lack of significant impact in either national markets or LRP activity zones indicates that CRS met USDA’s “do no harm” requirements. On the modest-to-moderate scale of the CRS’ local procurement, there was no discernible effect on beans, maize, or Incaparina prices. The evaluation team’s analyses of available data indicate that this project’s procurement did not have any likely impact on the prices of the affected commodities. Please see Annex 3 and the market impact chapter of the main body of the report for further detail. Please note that the only time series data available to the evaluators were at the national level rather than for source markets.


The evaluation team found that trade policies, including import tariffs, import quotas, subsidies, and exports, limited trade between the world market and domestic market, prevented integration with external markets. The team did find, however, that within the country maize markets are fairly integrated. Maize is highly protected, and tariff quotas are designed to maintain internal markets. CRS reported that as a result of the unusually high price of maize, on February 11, 2011, the Guatemalan Government, through Ministry Agreement 13-2011, authorized the importation of 82,000 MT of white maize for the following 90 days, in order to help stabilize consumer prices. Out of the nine companies that enrolled in the process, Maseca, Derivados del Maíz de Guatemala and Agroimportaciones S.A. were authorized to import 12,890 MT of maize. As of April 13, 2011, 9,635 MT of maize had been actually imported as a result of this announcement from the government. In 2010, just under 10,000 MT of commodities, including CSB, beans, rice, and vegetable oil, were shipped to Guatemala under a Title II emergency food aid program and close to 25,000 MT through a Title II development program, with an additional 9,750 MT of yellow corn provided through a Food for Progress


Reporting Requirement Notes grant and 12,000 MT of a range of commodities under a McGovern-Dole International Food for Education and Child Nutrition program.24 For 2011, 41,900 MT of commodities were programmed by USAID and USDA for shipment to Guatemala through Title II, the Food for Progress program and the McGovern-Dole International Food for Education and Child Nutrition program.25


1,224.47 MT of white maize, 159.47 MT of Incaparina, and 146.94 MT of beans were procured under the project.


Average total times, across 13 procurements, were as follows: • Maize took an average of 40 days total to procure, with 32 days used for

contracting and eight days for delivery; • Beans took an average of 39 days total, with 31 days for contracting and

eight days for delivery; and • Incaparina took an average of 46 days with 27 days for contracting and

19 days for delivery.


Maize commodity costs came in at an average of: $494.94 per MT, with an additional $49.49 per MT required for TSH. Beans commodity costs averaged $999.08 per MT over the four procurements, plus an additional $99.91 per MT for TSH. Incaparina cost $1,731.09 per MT on average for commodities, plus $173.10 per MT for TSH.

ii. Assess:


CRS reporting was found to be compliant with requirements. Sections on government interference and impacts on low-income consumers are rather sparse, but are included in the report.



Mayoreo Agrícola, a company located in Santa Rosa, markets agricultural inputs and grains. It supports small-scale farmer groups by allowing them to purchase seeds and other agricultural inputs on credit so that they can produce white maize, beans, and sesame seeds. Mayoreo Agrícola sold white maize and black beans to CRS at a higher price than that of the local market, which allowed the project to purchase from an increased number of small-scale farmers, albeit indirectly. In addition, in the 2011 agriculture production cycle, Mayoreo Agrícola recorded improved quality and packaging of the grains that they sell to their regular clients, as a direct result of this project’s quality requirements and the organization’s adherence to them under the project. Albay is a marketing and packaging company that sells rice and all types of beans at the national level. The company sold beans and maize to the project which gave them the opportunity to trade white maize for the first time. The evaluation team visited Agricoltor, a vendor that aggregates the produce of 400 farmers to provide for commercial and donor clients. Hurricane Agatha had ruined crops, which made it harder to amass the quantities needed for the LRP activity, but the vendor was able to source from outside the country to meet the

24 United States Agency for International Development and United States Department of Agriculture. U.S. International Food Assistance Report 2010. N.d. Web. 14 Apr. 2010. <http://www.usaid.gov/our_work/humanitarian_assistance/ffp/fy2010.ifarreport.pdf>. 25 US Food Aid Tables FY 2011,< http://www.fas.usda.gov/excredits/FoodAid/Reports/reports.html>.



quantities procured. The vendor reported capacity building from the special packaging requirements of the contracts with CRS. As a result of the project, the farmer’s association Asociación ACUS, which also sold maize and beans to CRS under the project, has formalized its provision of fertilizer and seeds to farmers, has defined their production areas in need of support, and has begun a marketing project. Farmers’ associations which submitted bids to participate in this project, but did not receive an award, began to analyze and improve their own production, marketing, and legal requirements to formalize their sales so that they would be in a better position the next time a similar competitive tender was issued. CRS further reported that small-scale farmers received training and technical assistance to improve their current agricultural production system and incorporate environmental conservation activities (e.g., establishment of forest tree and coffee nurseries, use of organic fertilizer and soil conservation practices to protect soil from erosion provoked by rain). The project also helped small farmers establish vegetable gardens with maize, radish, cucumber, carrot, coriander, and other crops to help families increase and diversify their availability of foods to improve their diet.

bb Markets

CRS reported that the price analyses conducted in the program found that the LRP activity had no statistically significant effects on retail prices in markets within the procurement and distribution zone and additionally that project-run LRP did not adversely affect normal market functioning in the geographic region within which the CRS activities took place. In the case of Incaparina, the LRP project did not affect supply or demand in the market as there is no serious supplier competition to provide Incaparina. This product is produced by only two companies. The evaluation team’s analysis found fully integrated maize markets that appeared not to suffer any price changes around the time of the LRP project procurements. For beans, a very small increase was seen in the national market around the November 2010 procurement of beans, but the quantity purchased was very small relative to market size. The evaluation team’s analyses of existing and available data indicate that this project was unlikely to have had an impact on the prices of the affected commodities. Please see Annex 3 and the market impact chapter of the main body of the report for further detail.


CRS reported that low-income consumers in participating communities benefited from the project through road rehabilitation and maintenance, improved facilities for schools and health posts, and reforestation and soil conservation activities. The evaluation team’s site visit confirmed the rehabilitation and maintenance work, which is a positive effect on community members who were not also project beneficiaries. Given the evaluation team’s findings of unlikely price effect, impacts on low income consumers are also not likely.

dd Program recipients The evaluation team spoke with project recipients in Guatemala, who mostly reported that they do grow small amounts of food on their land. Recent weather


Reporting Requirement Notes shocks (Hurricane Agatha) had resulted in low harvests. The team also visited several sites with gardens, school and road maintenance and grey water filters. Respondents were pleased with the food quality, saying it was even better than the quality of products found in their local markets. Respondents generally agreed that the food was delivered when they most needed it. Three respondents reported finishing the food prior to the next distribution (with the exception of Incaparina), which they attributed to large family size. Recipients preferred commodities distributed by the LRP project, especially maize and beans, as these are main staples in the Guatemalan diet. The fortified flour was also well-accepted by beneficiaries, as Incaparina provides essential proteins and nutrients to young children and conforms to the children’s tastes. Incaparina intake improved the health of the entire family, especially children and the elderly, as it helped them to return to a healthy weight. During the evaluation site visit, the team heard beneficiaries’ concerns that the end of the project was near, but the families still suffered from the effects of climate shocks – low harvests, high prices – that spurred the project. Several community members – while praising the project – also expressed dismay that the benefits were ending when the need still remained.

III Time of delivery Average total times, across 13 procurements, were as follows: Maize took 40 days total to procure, with 32 days required for contracting and eight days for delivery; Beans took a total of 39 days to procure, with 31 days needed for contracting and eight days for delivery; and Incaparina took an average of 46 days total to procure, with 27 days for contracting and 19 days for delivery.

iii. Compare: IV Quality and safety assurances Commodities delivered to the CRS’ warehouse were sampled by a regionally-

based governmental commodity inspector hired by CRS. The inspector conducted laboratory analysis to assure good quality, including measures of humidity level of the grains and aflatoxin testing. Aflatoxin testing found only very low levels, well within standards established by the Guatemalan Standard Commission. About 150 subsamples for every 3,000 quintales, or around 150 MT, of white maize received were taken randomly and sent for analysis. Beans underwent the same inspection process, only with fewer samples taken, as the quantity of beans purchased was lower than that of maize. The samples of fortified flour were analyzed once in order to compare the results to the manufacturer’s own quality testing.


TABLE 1: HISTORIC MAIZE PRODUCTION

Agricultural Cycle26 Harvested Area (In Manzanas27) Production (Quintals)28 Maize Yield

(Quintals per Manzana)

2004/05 860,000 28,233,900 32.83

2005/06 842,800 30,335,314 35.99

2006/07 825,944 32,840,155 39.76

2007/08 985,000 35,239,409 35.78

2008/09 1,224,600 37,954,987 30.99

2009/10 p/ 1,175,000 35,298,138 30.04

2010/11 e/ 1,200,000 35,827,610 29.86 p/ Preliminary data e/ Estimated data Source: USDA Local and Regional Procurement Pilot Project, Evaluation Report Final LRP Project Report Catholic Relief Services/ Guatemala

TABLE 2: HISTORIC BEAN PRODUCTION

Agricultural Cycle29 Harvested Area (Manzanas) Production (Quintales) Beans Yield (Quintales

per Manzana) 2004/05 312,493 4,108,320 13.15

2005/06 312,994 4,067,237 12.99

2006/07 315,498 4,148,581 13.15

2007/08 320,505 4,281,421 13.36

2008/09 328,518 4,401,221 13.40

2009/10 p/ 330,521 4,256,020 12.88

2010/11 e/ 330,521 4,319,821 13.07

Source: USDA Local and Regional Procurement Pilot Project, Evaluation Report Final LRP Project Report Catholic Relief Services/ Guatemala

TABLE 3: MAIZE CONSUMPTION (APPARENT) 2008 TO 2010

Year Quintals

Production Imports Exports Apparent Consumption

2008 31,715,468 1,281,841 90,263 32,907,046

2009 34,159,500 431,198 264,062 34,326,635

2009 31,768,290 861,847 47,475 32,582,662

2010 32,244,849 545,538 46,904 32,743,483


26 From May from one year to April of the next year 27 A manzana is equal to 0.70 hectares (7,000m2 or about 1.73 acres). 28 A quintal is equal to 100 pounds and is the most common measure for commodities in Guatemala. 29 Agriculture Cycle: from May of one year to April of the next year


TABLE 4: BEAN CONSUMPTION (APPARENT) 2004 TO 2010

Year Production (quintals)

Imports (quintals)

Exports (quintals)

Apparent Consumption (quintals)

2004 4,108,320 108,559 69,008 4,147,870

2005 4,067,237 105,357 24,113 4,148,480

2006 4,148,581 254,580 1,230 4,401,932

2007 4,281,421 175,836 50,840 4,406,417

2008 4,401,221 113,509 48,217 4,466,513

2009 4,256,020 179,335 10,540 4,424,815

2010 4,319,821 262,657 27,477 4,555,001


TABLE 5: AVERAGE PRICE OF WHITE MAIZE (IN QUETZALES PER QUINTAL)

Year Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

2005 91.08 94.42 93.82 88.92 88.31 94.46 96.77 93.50 84.08 91.69 79.38 75.00

2006 77.54 78.83 92.38 89.78 84.92 95.85 98.23 99.69 94.83 87.50 88.08 90.00

2007 104.77 116.83 120.08 115.70 116.15 128.23 137.69 137.38 128.45 109.64 95.42 90.00

2008 98.77 104.08 110.50 109.65 112.05 127.55 126.87 123.90 132.76 129.05 120.32 121.18

2009 127.65 128.90 135.50 134.95 130.63 132.52 139.04 138.33 127.33 108.71 102.76 102.65

2010 118.33 115.45 123.59 123.40 119.14 119.05 131.14 132.09 136.76 117.15 114.90 118.00

2011 140.10 157.00 175.04 173.89 - - - - - - - -

Source: Agri-Livestock Statistics Section-Ministry of Agriculture, in USDA Local and Regional Procurement Pilot Project, Evaluation Report Final LRP Project Report Catholic Relief Services/ Guatemala. Figures as of April 13, 2011

TABLE 6: PRICES OF BEANS AT THE NATIONAL MARKET (IN QUETZALES PER QUINTAL)

Year Months

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

2007 258.08 249.17 257.31 257.00 250.77 263.08 260.00 262.31 272.73 269.64 292.50 281.11

2008 272.31 281.54 281.43 313.50 372.86 448.75 492.17 500.75 460.24 479.52 487.37 432.65

2009 425.50 436.25 396.36 410.00 429.47 436.19 445.22 427.62 431.43 430.00 438.33 421.47

2010 410.56 402.00 382.27 371.50 358.57 368.10 360.00 351.82 337.14 382.25 386.19 398.75

2011 417.86 405.00 376.74 362.22 - - - - - - - -

Source: Agri-Livestock Statistics Section-Ministry of Agriculture as of April 13, 2011, in USDA Local and Regional Procurement Pilot Project, Evaluation Report Final LRP Project Report Catholic Relief Services/ Guatemala


TABLE 7: PRICES OF FORTIFIED FLOUR (INCAPARINA)

2003 2004 2005 2006 2007 2008 2009 2010 2011

Bag of 450 g of Incaparina Q3.72 Q3.91 Q3.79 Q3.93 Q4.33 Q5.44 Q5.69 Q5.96 Q6.53

Source: USDA Local and Regional Procurement Pilot Project, Evaluation Report Final LRP Project Report Catholic Relief Services/ Guatemala, (Q = Quetzales)

FIGURE1: RETAIL MAIZE PRICES, 2007-2011

Source: Annual Report of FAO projects in regions of Guatemala. 2011

FIGURE 2: RETAIL BLACK BEAN PRICES, 2007-2011

Source: Annual Report of FAO projects in regions of Guatemala. 2011


KENYA

WORLD VISION LOCAL PROCUREMENT PROJECT, EMERGENCY World Vision’s Project for Emergency Assistance in Kenya (PEAK) was implemented in Moyale, a district in the food deficient semi-arid region of northern Kenya. Most residents of Moyale are pastoralists and dependent on livestock for their livelihoods and food security. Moyale experienced emergency levels of acute malnutrition in 2011 as a result of a severe drought. Because of the drought conditions and the mainly pastoralist livelihoods of the Moyale people, very little of the food available in Moyale market is produced locally. The District Agriculture Office puts the food coming from Ethiopia at about 50 percent of the total food consumed in Moyale, food from Nairobi at 40 percent, and local production at only ten percent. The PEAK project procured 204 MT of beans, 245 MT of CSB, 1,144 MT of maize, 10 MT of salt and 82 MT of vegetable oil for 3,356 beneficiary households; all commodities were procured within Kenya via hard tender and distributed directly to beneficiaries.



Although Kenya is within the top 20 producers of maize globally, its yearly national demand of around 3.2 million MT means that the average yearly production of 2.75 million MT is insufficient to meet domestic demand. As such Kenya remains a net importer of maize. Kenya has also consistently been within the top-10 global producers of beans, but domestic consumption here too outstrips domestic supply, leaving Kenya a net importer. Concerted efforts to increase bean production to meet domestic demand have led to increases in cropping areas and yield since 2009, but production still falls short of demand. Over the past five years, the world has experienced two periods of rapid global food price increases. In late 2007 to early 2008, prices for many commodities around the world increased to historic highs. In early 2011, global food prices again began to rise, sparking concern that the instability observed in 2007-2008 would be repeated. In Kenya, the global food price increases in 2011 were exacerbated locally by poor rains which, in some areas, led to the most severe drought in 60 years. National wholesale prices for maize and beans were collected from the Food and Agriculture Organization Global Information and Early Warning System (FAO/GIEWS). Wholesale prices in Nairobi were analyzed as the vendors chosen for procurement were Nairobi-based. As shown in Tables 4, 5, and 6 below, maize prices held fairly steady year-round until the global food price crisis of 2007-2008. Beginning in February 2008, maize prices began to rise and remained high throughout 2009. In May 2009, maize prices reached $442 per MT. After the rainy season in late 2009, the harvest began, with maize prices finally beginning to drop and eventually falling to a five-year low of $174 per MT in July 2010. Prices returned to the 2006-2007 levels, but then began increasing in early 2011 and continued rising to an all-time high of $513 per MT in July 2011. In one year, maize prices increased 195 percent.30 Wholesale bean prices reacted more markedly to the 2007-2008 global food price crisis than maize. From February to May 2008, bean prices increased by nearly 60 percent. Prices stabilized somewhat in late 2008 and early 2009, but never returned

30 USDA Local and Regional Procurement Pilot Project, Evaluation Report –Final LRP Project Report World Vision/Kenya PEAK (Project for Emergency Assistance in Kenya). Unpublished. Nairobi, Kenya: World Vision/Kenya Country Office, 2011.


Reporting Requirement Notes to 2006-2007 levels. In March 2011, prices began to rapidly increase at a near identical rate to 2008. Prices peaked in June 2011 at $947 per MT, just below the May 2009 record high of $949 per MT. The Rift Valley is the major Kenyan production area for both maize and beans; Nakuru is the major market for the region. Table 6 illustrates historical price data (FAO/GIEWS) over the past three years for both commodities. Nakuru source market prices are lower than those in Nairobi, but follow the same patterns, indicating moderate levels of market integration. As with the Nairobi prices, the highest levels for beans in Nakuru are found in 2008 and 2011 and the highest levels for maize are found in 2009 and 2011. Because of the location of Moyale (the LRP project site), many of the crops brought to market come from neighboring Ethiopia. Table 7 shows the wholesale price of maize in Addis Ababa, the largest wholesale market in Ethiopia, and retail prices of maize and beans in Awasa, a wholesale and retail market which supplies Moyale. As in Kenya, Ethiopia saw high prices in 2008, though later in the year than in Kenya, and also in 2011, paralleling price rises in Kenya. Maize prices in Ethiopia increased in 2011, though not to the levels of 2008 or compared to those in Kenya. Bean prices increased in 2011 to levels surpassing 2008. Still, overall prices in Ethiopian source markets remained lower than those in Kenya. PEAK’s targeted beneficiaries were residents of the most rural areas of Moyale and had very remote access to the central market. Transportation infrastructure clearly played a role in constraining market integration between rural areas and central markets in this circumstance, although the country on the whole has relatively well integrated markets as explained above. More information on Nairobi source market prices can be found in Table 1 below.


All purchases were carried out using competitive tenders. For the first tender, World Vision issued an open tender on February 10, 2011, that closed on February 24, 2011. Thirty-one bids were received and World Vision chose to issue a contract to M/S Export Trading Company Limited for maize, pulses, CSB and salt, which included multiple delivery dates. World Vision also awarded a contract to M/S Techno Relief Services for vegetable oil. Both firms won the awards as they were able to meet the technical tender requirements and had the lowest prices. As described earlier, there were dramatic price increases in maize and beans beginning in April. This price increase spurred Export Trading, the main supplier from the first tender, to issue a notice that if World Vision could not renegotiate prices, then they would not be able to provide the deliveries of maize and beans remaining under the contract. World Vision conducted a price analysis which found that the revised prices were still below market, and in order to not disrupt distribution, an addendum to the contract was signed in June to readjust prices for July, August and September deliveries. In the original proposal for PEAK, 15 percent of the maize needed for the project was to come from farmers’ associations in Moyale. Because of the lack of rain during both the short and long rainy seasons, no surplus production existed, and none could be procured. Consequently, the drought and failed harvest increased the caseload of malnourished children. PEAK was requested to take on 200


Reporting Requirement Notes additional households with malnourished children for the months of July, August, and September; these additional beneficiaries and the need to replace local maize required an additional procurement to be made. On July 27, 2011, World Vision sent a request for bids to ten suppliers who had qualified from the previous tender. Export Trading issued the most competitive bid. A new contract for the remaining commodities was signed on August 9, 2011.


No evidence was found of any impact on producer or consumer prices from datasets provided by World Vision. The evaluation team’s analyses of existing and available data indicate that this project did not have a likely impact on the prices of the affected commodities, while Cornell University’s analysis found temporary market price increases for maize only. Please see Annex 3 and the market impact chapter of the main body of the report for further detail.


World Vision reported that there were no changes in Kenyan governmental policies or regulations during the time of PEAK implementation and therefore there was nothing which would have caused a shock to the market or which would have been considered interference, but rather they encountered a standard operating environment. USAID announced a major expansion of its global food aid prepositioning system in the fall of 2010. USAID awarded contracts for six new prepositioning sites, including one in Kenya. Note, however, that only some prepositioning stocks are intended for the country where the prepositioning warehouse is located, thus no supply shocks would result from the construction of this new prepositioning depot. USAID provided $101,975,000.00 in Title II emergency program funding for 109,840 MT of cornmeal, CSB, sorghum, vegetable oil, wheat flour, and yellow split peas in 2010; the USDA-funded McGovern Dole International Food for Education and Child Nutrition program provided 11,380 MT of vegetable oils, bulgur wheat, yellow peas, and CSB for 700 school children; and USAID provided an additional $706,600 from the Bill Emerson Humanitarian Trust to WFP for its emergency food aid program.31 USAID programmed 103,800 MT under Title II and USDA programmed 9,600 MT under the McGovern-Dole International Food for Education and Child Nutrition program to be shipped for Kenya during 2011.32


PEAK procured 1,144 MT of maize, 245 MT of CSB, 10 MT of salt, 204 MT of beans, and 82 MT of vegetable oil through the project.


Average procurement times were as follows: • Maize took 59 days total to procure on average, with 30 days needed for

contracting and 29 days for delivery. • CSB required 54 days overall to procure, with 30 days for contracting and

24 days for delivery. • Salt took 46 days total to procure, with 46 days for contracting. World

Vision took procurement of the commodities on the day of contract signature, so there was no additional delivery phase time.

• Beans took 51 days overall to procure, with 30 days needed to contract and 21 days to deliver.

• Vegetable oil required 49 days overall to procure, with 30 days for contracting and 19 days for delivery.

31 United States Agency for International Development and United States Department of Agriculture. U.S. International Food Assistance Report 2010. N.d. Web. 14 Apr. 2010. <http://www.usaid.gov/our_work/humanitarian_assistance/ffp/fy2010.ifarreport.pdf>. 32 U.S. Food Aid Tables FY 2011 < http://www.fas.usda.gov/excredits/FoodAid/Reports/reports.html>..


Reporting Requirement Notes VI Total cost of procurement,

including storage, handling, transportation and administrative costs

In terms of commodity costs, maize cost $313 per MT, beans cost $570 per MT, vegetable oil cost $1,593 per MT, CSB cost $674 per MT, and salt cost $230 per MT. Transport costs were charged by kilogram regardless of commodity. Storage and handling costs also were indiscriminant of commodity. Thus, total TSH for all commodity types amounted to $125 per MT.

ii. Assess:


World Vision reporting was found to be compliant with requirements.



The project had no negative impacts among producers since producers were constrained primarily by weather in failing to create adequate supply, something entirely out of the project’s domain. A complete lack of rainfall which contributed heavily to widespread drought in the lead up to the Horn of Africa crisis instead had the strongest impact on the restricted supply of commodities. The Horn of Africa drought affected 3.7 million Kenyans in the arid and semi-arid areas of PEAK project implementation. Any slight change in the price observed within the project period was ascribed by the traders to high transportation costs, drought, increasing of food prices due to low crop production and high fuel prices, but not the procurement modality. In World Vision’s project design, PEAK was to work with 20 local farmers’ associations, and provide training in improved techniques to produce, store, and package emergency food rations. However, Moyale district is one of the areas in Kenya that was hit by one of worst drought that affected the Horn of Africa region in decades. Due to the circumstances, the agriculture objective was removed by amendment. The participant in this circumstance had to change procurement approaches from partial soft tendering with farmers’ associations, to hard tendering, when the lack of rain affected smallholders’ production. Hard tendering arrangements were simple to modify and more guaranteed to meet the agreed delivery schedule.

bb Markets

World Vision reported that no shops owned by local traders were closed and normal business practices were not interrupted due to food distribution in local communities. World Vision reported no market disruptions due to food distributions. Cornell University analyzed the price impact of World Vision’s local procurement project on behalf of the Learning Alliance for this project using an econometric analysis. It found a 10 percent increase in the price of maize possibly resulting from the LRP project, but noted that other market disruptions may be behind these increases; Cornell’s analysis found no such price impacts for WFP’s purchases in the same time frame, and those purchases were 40 times larger by weight. The evaluation team found no likely evidence of market impacts of LRP project activities. For more detail, please see Annex 3 and the market impact section of the main body of the report.


Reporting Requirement Notes cc Low-income consumers World Vision explained that procurement and distribution of the above

commodities did not have any significant negative effect on market prices, nor did it displace non-beneficiary low-income consumers. As the evaluation team similarly found no evidence of price impacts, low income consumers likely felt no negative price effects from the LRP Project.

dd Program recipients World Vision reported that of the recipients interviewed, 96 percent felt they had more to eat, 46 percent spent less on staple foods, and 43 percent noted that children ate more often than before the project distributions began. Some 3,356 households were reported to have benefited from activities under the PEAK program, with average household size estimated at six individuals.

III Time of delivery Average procurement times are as follows: • Maize took 59 days total to procure on average, with 30 days needed for

contracting and 29 days for delivery; • CSB required 54 days overall to procure, with 30 days for contracting and

24 days for delivery; • Salt took 46 days total to procure, with 46 days for contracting and zero

days for delivery; • Beans took 51 days overall to procure, with 30 days needed to contract

and 21 days to deliver; and • Vegetable oil required 49 days overall to procure, with 30 days for

contracting and 19 days for delivery.

iii. Compare:


World Vision reported in the project’s final evaluation report that the procurement modality adopted was appropriate since the supplier was able to deliver the commodities in a reasonable time, had adequate storage facilities, and no prior history of commodity quality issues. World Vision submitted inspection reports from Bureau Veritas Kenya, Ltd., with their project closeout report. The evaluation team reviewed a sample of inspection reports. For CSB, Bureau Veritas analyzed components such as vitamins, niacin, and thiamine. They also examined color, texture, taste, and odor. For maize and beans, they determined the levels of moisture, pest damaged grains, foreign matter, live insects, etc. Testing for aflatoxin was conducted on maize samples. No quality and safety issues were identified.


TABLE 1: COMMODITY PROCUREMENT REPORTING

Date of Procurement

Commodity Type

Local Market Price (one week before

procurement)

Local Market Price (date of procurement)

Local Market Price

(one week after procurement)

Area and Country of Procurement

4/1/2011 Maize 287 292 321 Nairobi, Kenya

4/1/2011 Beans 606 705 733 Nairobi, Kenya










9/1/2011 Beans 657 700 794 Nairobi, Kenya *Prices in $/MT

Source: USDA Local and Regional Procurement Pilot Project, Evaluation Report –Final LRP Project Report World Vision/Kenya PEAK (Project for Emergency Assistance in Kenya)

TABLE 2 (FROM MSI EVALUATION DATASET)

Commodity Procurement

Quantity Average

Commodity Cost (USD/MT)

Average TSH Cost (USD/MT)

Total (Extended) Cost

Maize (6) 1,144 MT $313.00 $125.00 $501,072

CSB (7) 245 MT $1,593.00 $125.00 $420,910

Salt (6) 1.7 MT $230.00 $125.00 $86,975

Beans (6) 204 MT $570.00 $125.00 $141,780

Vegetable oil (4) 82 MT $674.00 $125.00 $65,518 Note: numbers in parentheses are numbers of procurements of that commodity


FIGURES 1-2: WHOLESALE MAIZE AND BEAN PRICES, 2006-2011, IN USD/MT


FIGURE 3: WHOLESALE MAIZE AND BEAN PRICES, NAKURU-RIFT VALLEY, KENYA, 2008-2011, IN USD/MT


Wholesale Maize Prices - Nairobi

$442

$174

$513

0

100

200

300

400

500

600

Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep

US$

/Met

ric T

on

2006/7

2007/8

2008/9

2009/10

2010/11

Wholesale Bean Prices - Nairobi

$453

$949 $947

0

100

200

300

400

500

600

700

800

900

1000

Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep

US$

/Met

ric T

on

2006/7

2007/8

2008/9

2009/10

2010/11

Wholesale Prices in Nakuru - Rift Valley

$833

$460

0

200

400

600

800

1000

1200

Jan-0

8

Mar-08

May-08

Jul-0

8

Sep-0

8

Nov-0

8

Jan-0

9

Mar-09

May-09

Jul-0

9

Sep-0

9

Nov-0

9

Jan-1

0

Mar-10

May-10

Jul-1

0

Sep-1

0

Nov-1

0

Jan-1

1

Mar-11

May-11

Jul-1

1

Sep-1

1

US

$/M

etri

c To

n

Beans Maize


FIGURE 4: WHOLESALE AND RETAIL MAIZE AND BEAN PRICES, 2008-2011, IN USD/MT


Prices for Awasa (Retail) and Addis Ababa (Wholesale)- Ethiopia

$840

$940

$599

0

100

200

300

400

500

600

700

800

900

1000

Sep-07

Dec-07

Mar-08

Jun-0

8

Sep-08

Dec-08

Mar-09

Jun-0

9

Sep-09

Dec-09

Mar-10

Jun-1

0

Sep-10

Dec-10

Mar-11

Jun-1

1

US$

/Met

ric T

on

Awasa - Maize Awasa - BeansAddis Ababa - Maize WS


MALAWI

WFP LOCAL PROCUREMENT PROJECT, DEVELOPMENT The WFP project in Malawi purchased roughly 2,400 MT of maize, pulses, and CSB Plus in Malawi through the Purchase for Progress (P4P) pilot during the 2010-2011 agricultural marketing season. These purchases were financed through a $1.7 million donation from the USDA LRP Project. P4P is a program that utilizes WFP procurement to create incentives for farmers to increase productivity while more efficiently linking buyers and sellers to markets. Using a range of procurement modalities and tendering practices, food for WFP’s humanitarian operations was purchased in ways that were intended to enhance development outcomes for smallholder farmers; this engagement was designed to build the capacity of smallholders to participate in profitable market relationships and thereby increase incomes and livelihoods.



WFP reported a very strong bumper crop in 2009 in Malawi of 3.4 million MT of maize, and 3.9 million MT in 2010. Bumper crops in Malawi, coupled with similarly strong harvests regionally, resulted in excess supply in Southern Africa and led to more stable, less pronounced seasonal price changes across the country and region. The increasing supply of maize over the last ten years is shown in Figure 1. WFP’s baseline report provided time series data indicating a clear upward trend in prices for most markets (maize prices are shown in Figures 2 and 3). The upward price trend is clearest for nominal prices. Additionally, real prices, adjusted using the national food consumer price index, trend upward significantly in most markets. Over time, there appeared to be a structural component of markets that caused real prices to increase. While the food price crisis in 2008 did have a statistically significant impact on the upward trend in real prices, the real price rises remain significant even when excluding 2008. Government of Malawi interventions in the market (e.g. trade bans in 2008 and price floors in 2008, 2009, and 2010) deterred some productive private sector investment in areas such as storage and transportation, which has resulted in increasing costs. WFP’s baseline report also illustrates an example of price fixing by large agricultural traders, behavior which carries reverberations for most all producers and consumers in the marketplace. WFP reports that collusion between small vendors/farmers and small traders has occurred on occasion when small traders are located within close proximity to each other and have the opportunity to cartelize. More importantly though, WFP’s baseline report echoes findings from other projects in central Africa: large trader price fixing. However, the report also notes that in some cases, a large trader with enough market share in a particular market is able to force other market actors’ prices artificially downward, since medium and small traders are reliant on the large trader to sell their commodities. WFP notes that these large trader price fixing monopolistic tendencies are most often witnessed in rural markets without easy access to major roadways. This limits market integration due to infrastructure issues and market actors must use the nearest markets regardless of price.


Reporting Requirement Notes I Extent of competition for

procurement bids Commodities were procured through three different approaches, including direct purchases, hard tenders and soft tenders. Ten direct purchases of white maize totaled 863.2 MT. Direct purchases were made only with FOs throughout six districts of Malawi. Hard tender procurements using the Agricultural Commodity Exchange for Africa (ACE) had greater competition and provided 765 MT of pulses in total, with four to seven bidders for each tender. Three soft tenders were also issued for cowpeas, but only one was completed, for 76 MT of cowpeas. This soft tender received only one bid.


WFP reported that across target P4P districts, wholesale maize prices remained equally or more stable than retail prices, as measured by a coefficient of variation. The lack of pronounced volatility in these prices provides additional support for the assertion that P4P purchases did not affect local supply and demand balances enough to noticeably influence market prices. WFP reported on maize procured via direct purchases with smallholder farmer organizations in six districts during the period from August to December 2010. In all districts, retail maize prices at the organizations’ primary local markets remained below the five-year average plus one standard deviation. In five of the six districts, the retail prices in both local and secondary markets remained more stable from August to December 2010 than did the five-year averages observed historically. These findings suggest no negative retail price impact arising from WFP purchases. WFP reported that cowpea prices were more volatile over the project period than were historical data on bean prices. WFP suggested that it is unlikely that this price volatility resulted from a 76 MT purchase, considering the total levels of beans available. Furthermore, the additional 765 MT of pulses purchased through ACE were supplied over a two month delivery period. WFP estimates that this likely constitutes a relatively small share of national marketable supplies. WFP reported that retail prices remained abnormally low compared to previous years and below the price floor. In the few districts where retail prices did eventually reach the price floor set by the Government of Malawi, this did not occur until December 2010 when prices typically approach their seasonal peaks. This indicates strong aggregate supply smoothing seasonal price behavior, which combined with small quantities procured by P4P, left retail markets accessible to low-income consumers. WFP reported one occurrence in Mzimba district where a sudden retail price spike directly preceded delivery of 40 MT of maize from Manyamula Legume Producers Co-operative. The 40 MT delivered represented a 60 MT default from the 100 MT originally contracted. It is possible that this purchase may have impacted the market, although maize prices remained below the Government of Malawi’s price floor. With regard to CSB, WFP reported that in the absence of historical data, analysis of CSB purchases considers the price behavior of the principal ingredients: maize and soy. Retail maize prices over the procurement period were more stable than the previous five years in three major maize trading markets. Soy also demonstrated stable retail prices over the procurement period.


Reporting Requirement Notes The evaluation team’s analyses of existing and available data indicate that this project did not have any likely impact on the prices of the affected commodities. See Annex 3 and the market impact chapter of the report for further detail.


In 2006, the Government of Malawi began subsidizing fertilizers, contributing to a tripling of corn production over three years. This program, known as the Farm Input Subsidy Program (FISP), also provides seeds, pesticides and top soil to targeted poor smallholder farmers. Although the program has been quite successful in increasing output, it has been accused of non-transparent procurement procedures, poor planning and low cost-effectiveness. In 2008, the Government of Malawi’s nationalized the purchase and sale of maize in Malawi, giving sole price-setting rights to the parastatal Agriculture Development and Marketing Corporation (ADMARC). ADMARC raised the prices of 50kg maize bags from $12 to $18 in late 2011. The government, through ADMARC, regulates the price of maize in an effort to safeguard against excessive demand/supply. Since 2006, consistent surpluses of corn have, at times, oversaturated the market, driving down prices and discouraging smallholder farmers in subsequent seasons. For this reason, ADMARC raised prices to ensure continued productivity and profitability. To deal with domestic food shortages more efficiently, the Government of Malawi standardized the buying and selling of maize through ADMARC. Criticisms of ADMARC are that it buys maize at too low a price and that its stocks of maize run out too quickly. ADMARC consistently lacks adequate funds to buy sufficient quantities of maize. Buyers and sellers tend to adhere rather loosely to the prices set by ADMARC. ADMARC is supposed to buy from farmers, but recent fieldwork suggests that it buys most of its maize from traders. Government-imposed price floors have also limited the ability of the ACE to increase price transparency and discovery in the marketplace. U.S. food assistance activities in 2010 included 6,690 MT of CSB provided through the USDA-funded McGovern-Dole International Food for Education and Child Nutrition program, and a Title II development program that provided 17,190 MT of CSB, vegetable oil, wheat and beans.33 In FY2011, USAID programmed to ship 21,200 MT of commodities under Title II programming and USDA programmed to ship 7,800 MT of CSB under the McGovern-Dole International Food for Education and Child Nutrition program.34


WFP reported that in total, 863 MT of white maize were procured from smallholder farmers’ organizations and ten MT of white maize were procured from large traders through the ACE. 76 MT of cowpeas were procured from small traders via a soft tender and 765 MT of cowpeas and pigeon peas were procured from large traders through the ACE. 870 MT of CSB Plus were also procured from large traders through the ACE.


Average overall procurement times were as follows: • White maize took 162 days overall to procure, with 65 days for

contracting and 97 days for delivery. Please note: These times reflect the development activities of direct purchase from targeted FOs.

• CSB Plus took 63 days overall to procure, with 16 days for contracting and 48 days for delivery.

33 United States Agency for International Development and United States Department of Agriculture. U.S. International Food Assistance Report 2010. N.d. Web. 14 Apr. 2010. <http://www.usaid.gov/our_work/humanitarian_assistance/ffp/fy2010.ifarreport.pdf>. 34 U.S. Food Aid Tables FY2011, < http://www.fas.usda.gov/excredits/FoodAid/Reports/reports.html>.


Reporting Requirement Notes • Pulses required 53 days to procure, with 19 days for contracting and 34

days for delivery.


Average commodity costs were as follows: Direct purchase: White maize cost $230.43 per MT. Soft Tender: Cowpeas cost $471.13 per MT. Hard tender:

• White maize cost $212.20 per MT; • Cowpeas cost $449.66 per MT; • Pigeon peas cost $464.19 per MT; and • CSB Plus cost $436.19 per MT.

TSH costs were based on weight evenly and averaged $131.31 per MT for each type of commodity procured.


this section have been met WFP reporting was found to be compliant with requirements. However, historic supply and price information focused on maize. No historic demand information was found for any commodities; moreover, no specific information was found on the impact of the program on targeted food aid recipients.

II Local and regional agricultural producers, including large and small agricultural producers

Farmers reported earning more by selling to WFP than their counterparts earned by selling to traders or small vendors. Farmers selling to WFP also indicated that being paid in one lump sum for a significant portion of their marketable surplus was preferable to repeated smaller sales throughout the marketing season that are typical of marketing through small traders. Many project-supported FOs reported an increase in the enrolment of new members after they observed the WFP buying process the previous season. FOs reported that successful marketing through WFP also contributed to better relationships and communication between FO management and members. Despite this positive development, member contributions from within FOs were also reported to be highly unequal. Only a few farmers contributed to each WFP contract. Those that were able or willing to contribute were generally characterized by one or a combination of the following factors: more willingness to take a risk on an unknown buyer (WFP); having sufficient surplus maize; having received timely information from contacts in FO management; willing to meet WFP quality specifications; and having sufficient funds to buy new bags as required. Factors that weighed into the decisions of some farmers not to participate included an inability to pay organization membership fees, the need for quick payment after harvest to smooth household consumption and repay loans, and a desire to wait and observe the success of initial WFP contracts. Focus groups did indicate that the quantities demanded by WFP have incentivized participation in FOs. While women in some FOs have been directly targeted for input subsidies and trainings by government and NGOs, FOs reported that active female participation remained limited. Low participation by women was attributed in part to traditional gender roles: marketing was said to be considered a men’s activity.


Reporting Requirement Notes WFP held focus group discussions where participants reported having learned how to meet quality standards and improve production. Successive purchases in the final two years of P4P implementation should solidify these gains and continue to build the confidence and capacity of smallholder farmers to engage in sales to institutional, quality-oriented buyers.

bb Markets

WFP reported that paying above the average market price could have been distortive in different supply conditions, but with strong maize supply during the LRP activities, WFP reported that the practice resulted in no price distortion. WFP further reported that the price trend analysis does not indicate that the higher price paid by WFP resulted in additional price inflation or volatility that impacted low-income consumers. WFP explained that maize markets were very poorly integrated with world markets. Malawi markets generally have higher prices than international markets and more volatility. International maize markets were relatively stable through 2007, a trend which did not translate to Malawian markets, though increasing prices and volatility in the world market did affect Malawi markets in 2008. It seems that price volatility can easily translate from international markets to Malawi but the same cannot be said for price stability. The evaluation team’s analyses of existing and available data indicate that this project did not have any likely impact on the prices of the affected commodities. See Annex 3 and the market impact chapter of the report for further detail.

cc Low-income consumers WFP reports that P4P procured small quantities of commodities deliberately so as to leave retail markets open and accessible to low-income consumers. Moreover, WFP reports that during the P4P procurement period, local and national wholesale and retail prices were already favorable to low-income consumers as there was significant surplus production in agricultural markets. No price inflation or volatility that could have affected low-income consumers was reported. The evaluation team finds no likely price impacts of the LRP project in Malawi, which makes negative impacts on low-income consumers also unlikely. For further detail on the market analysis, see Annex 3 and the market impact section of the main body of the report.

dd Program recipients WFP distributed these commodities to beneficiaries through their Protracted Relief and Recovery Operation (PRRO) in Malawi. Beneficiaries included households affected by drought, households caring for orphans and vulnerable children, chronically ill people, and families with malnourished children under five years of age. In total, over 400,000 individuals received rations under this project. In addition to providing food aid to beneficiaries, this project targeted support and capacity building for FOs in food storage, quality standards, contracting and management.

III Time of delivery Average overall procurement times were as follows: • White maize took 162 days overall to procure, with 65 days for

contracting and 97 days for delivery. Please note: These times reflect the development activities of direct purchase from targeted FOs.

• CSB Plus took 63 days overall to procure, with 16 days for contracting and 48 days for delivery.

• Pulses required 53 days to procure, with 19 days for contracting and 34 days for delivery.


Reporting Requirement Notes iii. Compare:

IV Quality and safety assurances WFP reported that the suppliers of food commodities are required to confirm in their bids that the quality of the commodity they are offering satisfies WFP’s quality specifications as stipulated in the tender document. After the contracted supplier reported the readiness for delivering the consignment, WFP contracted with SGS, an inspection company, to collect a sample of the commodity for laboratory analysis (including aflatoxin testing). The sampling method and analysis technologies are based on international standards. The WFP food control unit periodically audited the laboratory and these audits confirmed compliance to required standards. Lastly, the quality of food is double-checked by a superintendent upon delivery, through a visual inspection of ten percent of the delivered quantity. WFP reported that it did experience some quality problems such as weevil infestation. WFP’s quality control unit advises in such cases whether deviation is within acceptable tolerance levels, or must be addressed through additional treatment.

FIGURE 1: MAIZE PRODUCTION IN MALAWI FROM YEARS 1963 THROUGH 2012

Source: http://www.indexmundi.com/agriculture/?country=mw&commodity=corn&graph=production

0

500

1000

1500

2000

2500

3000

3500

4000

1961

1963

1965

1967

1969

1971

1973

1975

1977

1979

1981

1983

1985

1987

1989

1991

1993

1995

1997

1999

2001

2003

2005

2007

2009

2011


FIGURE 2: MAIZE PRICES IN MALAWI SINCE 1992

Source: http://www.indexmundi.com/commodities/?commodity=corn

FIGURE 3: NATIONAL MAIZE PRICES RELATIVE TO LRP TONNAGE, 2004 – PRESENT

Source: World Food Programme. “Baseline Markets Report and Monitoring Strategy”. Malawi Country Office. July 2010.


MALI

CRS LOCAL PROCUREMENT PROJECT, DEVELOPMENT

CRS implemented a school feeding project in the Mopti region of Mali. Total funding for the project was $106,098 so it was a small scale project, procuring only 31.45 MT of millet, 7.01 MT of rice, and 7.67 MT of cowpeas. All commodities were procured and distributed in Mali. The commodities were procured through voucher fairs and competitive hard tendering. The commodities acquired through hard tendering were delivered directly to schools, whereas at the voucher fairs, school management committees were responsible not only for the set up and execution of the fairs, but also for the transport of commodities back to schools, and the commodities’ storage. Farmers acted as suppliers at voucher fairs, and traders supplied goods under hard tendering.



Historical market production of millet is relatively stable. Historical market production of rice has more than doubled from 2006-2007 to 2010-2011, as shown in Table 1. No millet imports were reported. The FAO reports that cowpeas are neither among Mali’s top 20 crops produced nor among Mali’s top 20 agriculture product imports. The evaluation team reported the secular trend in wholesale cereal prices in Bamako (Figure 1), corrected for general price inflation (as measured by the consumer price index, CPI). The rice market in Mali, with strong links to the world market, operates largely independently of the market for other cereals for which rice is an imperfect substitute. The price of rice is generally about twice as high as the price of millet. Despite the noticeable 2008 spike in rice prices corresponding to a period of high global food prices, real price levels of rice and millet in Mali remained fairly stable over the January 2006 to January 2011 period. The deflated millet prices vary significantly from year to year, in both level and profile (Figure 2). In particular, price levels were significantly higher than normal following the poor 2009 rains. On average, prices peak in July and August at the end of the “hungry season” as the rains start, and fall through the fourth-quarter harvest (and immediate post-harvest) period, reaching their nadir in December. Prices thereafter tend to rise until the next hungry season. No historic data were found for market price movements of cowpeas, either in the project final report or in external secondary sources.


CRS reported that 10 out of 12 project schools used voucher fairs to purchase commodities. While the level of competition at the fairs is not clear, CRS reported the process as transparent and successful, pre-selecting farmers on their ability to supply the required quantity at specified quality standards. One school in the Djenné district used a call for bids to purchase food for their canteen. Commodity needs for this school were divided into three batches to be fulfilled by three different producers. In all, six bids were received. Another project school in the village of Taga had wanted to host a voucher fair but was not able to do so because the quantity required could not be procured due to low quality of available stocks. The Taga school thus opted to switch to a hard tender for cowpeas in cooperation with that of the tender put out by another school.

II Impact of the procurement of the eligible commodity on

CRS reported that the LRP activity did not have any effect—positive or negative—on commodity prices in Mali during the period of the activity. The


Reporting Requirement Notes producer and consumer prices in the market

small size of the LRP activity, amounting only to $106,000 of market activity negated the likelihood of any impact. The evaluation team’s analyses of existing and available data indicate that this project had no likely impact on the prices of the affected commodities. Please see Annex 3 and the market impact chapter of the main body of the report for further detail.


The evaluation team found that the Government of Mali has, at times, formally banned cereal exports during periods of food shortages. The goal was to halt further price rises by stemming cereal outflows. Implementation at border posts may be strict but many cargoes can circumvent border posts if the incentives are big enough. In some cases, customs officers enforce bans for exports to one country but not to another. To the extent that bans effectively lower prices within Mali, they limit the incentives to farmers to produce more cereals.35 U.S. food aid shipments to Mali included, in 2010, approximately 6,500 MT of bulgur wheat, vegetable oils, CSB, and green split peas under Title II, and over 30,000 MT of wheat and soybean meal under a CCC-funded Food for Progress grant.36 In FY2011, USAID programmed to ship 2,300 MT of commodities under Title II and USDA programmed to ship 5,000 MT under the Food for Progress program.37


0.98 MT of cowpeas and 4.87 MT of rice were procured using hard tenders while 31.45 MT of millet, 6.69 MT of cowpeas, and 2.14 MT of rice were procured through a voucher program.


For the cowpeas and millet procured through hard tenders, the average procurement time overall was 64 days with 24 for contracting and 44 days for delivery. For the voucher program, the average time from initial discussions with supplying farmers through the first redemptions amounted to 50 days.


Average prices paid for commodities per MT were as follows: Hard tenders: Cowpeas cost $742.86 per MT and Rice cost $578.02 per MT. TSH costs across both hard tendered procurements amounted to $221.42 per MT. Voucher program: Millet costs were $332.15 per MT; Cowpeas costs were $577.60 per MT; and Rice costs were $638.95 per MT. TSH costs across all voucher purchases amounted to $95.30 per MT


this section have been met CRS reporting was found to be compliant with requirements; however no historic data were found for the market supply, demand, or price movements of cowpeas.

35 http://www.aec.msu.edu/fs2/srai/Agricultural_Globalization_in_Reverse_MSU_Crans-Montana_paper_final.pdf 36 United States Agency for International Development and United States Department of Agriculture. U.S. International Food Assistance Report 2010. N.d. Web. 14 Apr. 2010. <http://www.usaid.gov/our_work/humanitarian_assistance/ffp/fy2010.ifarreport.pdf>. 37 U.S. Food Aid Tables FY2011, < http://www.fas.usda.gov/excredits/FoodAid/Reports/reports.html>.


Reporting Requirement Notes II Impact of different methodologies and approaches on: aa Local and regional


No specific information was found for impacts on local large-scale agricultural producers, who were not a target of the program. CRS reported that small farmers were motivated to increase production volume, improve quality, and diversify their crops to increase their revenue as a direct result of this project. When interviewed by CRS staff, some farmers supplying to voucher fairs noted that even if they were not able to sell the entirety of their stock through the fairs, the remainder of the commodities would sell quickly at the market due to the high quality of the commodities. Cowpeas had long ceased to be a major commodity on the market in the implementation region, but CRS procurements stimulated increased production, an attitude which over time might have the potential to restructure the market. In order to meet the quality standards of the fairs, the post-harvest treatment of food products was done very carefully. No losses or decreases in revenue were reported by producers or any other actors involved in the implementation of the project.

bb Markets

CRS further reported no negative impacts on producers, markets, or consumers. CRS reported that the LRP activity did not have any effect—positive or negative—on commodity prices in Mali during the period of the activity. The small size of the LRP activity, amounting only to $106,000 of market activity, negated the likelihood of any impact. The evaluation team’s analyses of available data indicate that this project had no likely impact on the prices of the affected commodities. Please see Annex 3 and the market impact chapter of the report for further detail.


As part of the Learning Alliance, Cornell University analyzed the impact of the program on low-income consumers. Their report observed that voucher fairs operated independent of local markets, and therefore had no effect on low-income consumers. This was because farmers sold goods at the same price as they had before and immediately after the voucher fairs. As the evaluation team similarly finds no price impacts from the LRP project, low-income consumers likely experienced no price impacts from the voucher fairs.

dd Program recipients Local procurements made through this project provided meals to 2,892 students and 42 cooks over a period of about nine months, adding up to 93 meals for each individual over the project lifecycle. Direct distribution of commodities and the voucher program both provided incentives for children to attend school, as a result of the locally procured hot meal. School children preferred locally procured commodities over those purchased internationally, as locally procured commodities were more familiar to their taste and diet. Local procurement also allowed supplying farmers and traders to pay back their loans to local banks, injecting cash into these institutions which also lend to local women through microfinance programs. Working through school management committees also proved to be a very valuable investment in capacity building, as school management committees felt much more invested in the efficiency and success of the school feeding programs when the food was procured locally. CRS reported that both voucher and hard tendering procurement approaches reinforced the capacity of school management committees to purchase food and meet the needs of the school canteens using locally available resources. These methods improved the efficiency of the school purchasing process.


Reporting Requirement Notes III Time of delivery For the hard tendered cowpeas and millet, the average procurement time overall

was 64 days with 24 days for contracting and 44 days for delivery. For the voucher program, the average time from initial discussions with supplying farmers until first redemptions was 50 days.

iii. Compare: IV Quality and safety assurances At each voucher fair the local agricultural service conducted an inspection during

registration for each vendor. Humidity tests were performed using random sampling. When the quality control tests concluded, the local agricultural service delivered a certificate attesting that the products purchased met quality standards. Products not meeting quality standards were rejected. CRS reported that commodities brought to voucher fairs were consistently of higher quality than products sold on reference markets. As an additional quality control measure, CRS trained members of the school management committees in quality control practices, for oversight of cooking and proper hygiene for school cooks. Commodity Inspection Reports were provided by CRS. Commodities were tested by the Office des Produits Agricoles du Mali (OPAM) for damaged grains, foreign materials, and moisture content. The reports showed commodities tested to be acceptable.

TABLE 1: MALI NATIONAL SUPPLY AND CONSUMPTION OF LRP COMMODITIES

Commodity Attribute (1000 MT) 2006/2007 2007/2008 2008/2009 2009/2010 2010/2011

Millet Production 1,129 1,175 1,242 1,170 1,175 Imports NA NA NA NA NA Domestic

Consumption 1,129 1,175 1,242 1,170 1,175

Rice, Milled Production 695 714 873 1,043 1,523 Imports 105 100 100 165 125 Domestic

Consumption 800 814 958 1,137 1,634

Source: Mali LRP Final Report using USDA bullet points 6Oct11 (CRS Mali), in which the report cites “USDA Foreign Agricultural Service website”; further information not found on the website


FIGURE 1: WHOLESALE CEREAL PRICES IN BAMAKO, 2006 – 2011

Source: FAOSTAT Note: The vertical axis title “thousands of 2000 CFA francs per 100 kg sack” refers to CFA francs at their average value in the year 2000.

FIGURE 2: DEFLATED MILLET PRICES IN BAMAKO 2006 – 2010

Source: FAOSTAT. http://faostat.fao.org/ Note: The vertical axis title “thousands of 2000 CFA francs per 100 kg sack” refers to CFA francs at their average value in the year 2000.

0

10

20

30

40

Jan-06 Jan-07 Jan-08 Jan-09 Jan-10 Jan-11

thou

sand

s of

200

0 CF

A fr

ancs

per

100

kg sa

ck

rice millet

0

5

10

15

20


thou

sand

s of 2

000

CFA

fran

cs

per

100

-kg

sack

2006 2007 2008

2009 2010 MEAN


MALI

WFP LOCAL PROCUREMENT PROJECT, DEVELOPMENT

WFP’s P4P project in Mali procured 2,251.45 MT of millet, 504.1 MT of sorghum, and 48.7 MT of cowpeas in fifty separate procurements between February 2010 and February 2011. The various contract dates for these procurements were from December 2009 through February 2011. All commodities were procured and distributed in Mali. The commodities were purchased through Forward Delivery Contracts (FDCs) and direct contracts, both different types of direct purchasing mechanism. The commodities were distributed directly to the beneficiaries. All commodities were procured from smallholder farmers’ unions, cooperatives, and associations participating in the project. Farmers’ unions are umbrella organizations that comprise farmers’ cooperatives. Cooperatives and associations are made up of farmer groups; cooperatives are officially registered as for-profit, while associations are non-profit. The commodities were distributed to 363,256 beneficiaries under WFP’s ongoing Country Program and PRRO.



While data for national historical market supply could not be found, WFP reported production of millet from Bankass (Figure 1) and sorghum from Koutiala (Figure 2) for the years 2008 to 2011. Bankass and Koutiala are reported as large areas of millet and sorghum production in Mali respectively. WFP did not report on the supply of cowpeas in the market. The evaluation team found that the prices of millet and sorghum, close substitutes, track each other tightly. Real prices for millet and sorghum have remained fairly stable over five years’ time from January 2006 to January 2011(Figure 3). The team further found that deflated prices for millet vary significantly from year to year, in both level and profile. In particular, price levels were significantly higher than normal following the poor 2009 rains. On average, prices peak in July and August at the end of the “hungry season” as the rains start, and fall through the fourth-quarter harvest and immediate post-harvest period, reaching their nadir in December. Prices thereafter tend to rise until the next hungry season (Figure 4). The FAO reports that cowpeas are neither among Mali’s top 20 crops produced nor among Mali’s top 20 agriculture product imports; as such, secondary data are not available on cowpea price movements.


WFP’s local procurements for the P4P program were made through hard tenders, via direct contracting and FDCs in which FOs made commitments prior to harvest to guarantee the delivery of certain quantities of stock after the harvest. FDCs benefit FOs in that they guarantee future purchases, encourage production and improve FOs’ familiarity with standard contracting procedures. Direct contracts were used to purchase the FOs’ excess stocks after the harvest. Both approaches were non-competitive, although the FOs were still required to meet quality standards in the provision of commodities.


Reporting Requirement Notes II Impact of the procurement

of the eligible commodity on producer and consumer prices in the market

WFP reported no significant price distortions during the procurement period that could be attributed to the LRP project. Market price data on millet and sorghum (collected before and after each procurement) support this conclusion. WFP did not collect market price information on cowpeas, thus no assessment can be made. The evaluation team also concluded from the available data that the procurements of millet and sorghum did not have any appreciable impact on the market prices of those commodities. Mean price data collected by WFP were found to vary only negligibly over the time period of the WFP LRP project, and mean procurement size was small – 70 MT on average. Regarding cowpeas, time series data for the price of cowpeas were not provided in the WFP report. The team found that, while any impacts on the prices of cowpeas from the LRP purchases are extremely unlikely, some impact on local market prices of cowpeas cannot be ruled out. For more detail, see Annex 3 and the market analysis section of the main body of the report.


The evaluation team found that the Government of Mali sometimes formally bans cereal exports during periods of food shortages. The goal is to halt further price rises by stemming cereal outflows. Implementation at border posts may be strict but many cargoes can circumvent border posts if the incentives are large enough. In some cases, customs officers enforce bans for exports to one country but not to another. To the extent that bans effectively lower prices within Mali, they limit the incentives to farmers to produce more cereals.38 WFP reported that to boost agricultural production the Government of Mali, in 2008, devised a rice production program with the objective of producing 10 million MT of cereal by 2012. Originally intended to boost rice production by providing input subsidies and fertilizer to farmers, the initiative has widened to include increased production of other cereals – maize, millet, and sorghum. At the end of 2010, gross production through this program was reported to have reached 4.7 million MT. WFP further reported that in response to persistently high global food prices and volatility risks to local markets, the Government of Mali has taken a number of measures designed to protect the purchasing power of low-income households. These include:

• Public and community stock rebuilding at harvest time; • Making supply available in deficit areas at affordable prices; • Improving the overall supply by offering stocks publicly at

reasonable prices during the lean season; and • Authorizing tariff-exempt imports, particularly for rice.

U.S. food aid shipments to Mali included, in 2010, approximately 6,500 MT of bulgur wheat, vegetable oils, CSB, and green split peas under a Title II

38 http://www.aec.msu.edu/fs2/srai/Agricultural_Globalization_in_Reverse_MSU_Crans-Montana_paper_final.pdf


Reporting Requirement Notes development program, and over 30,000 MT of wheat and soybean meal under a CCC-funded Food for Progress grant.39 In FY2011, USAID programmed to ship 2,300 MT of commodities under Title II and USDA programmed to ship 5,000 MT under the Food for Progress program.40


WFP purchased 2,201 MT of millet and 620 MT of sorghum through FDCs arrangements made in the first two years of the program. However, in the first year only 489 MT of millet were provided through the FDCs, with 504 MT of sorghum and 915 MT of millet provided in the second year. FDC defaults over the period amount to 36.2 percent for millet and 18.7 percent for sorghum. Producers’ ability to deliver grain to the P4P project in FDCs increased in the second year. WFP reported that FOs were increasingly committed to meeting their delivery targets when the FDCs were honored at market prices. WFP reported that direct contracts with FOs were used to fill in gaps where quantities of cereals were not fully acquired through FDCs. Purchasing of cowpeas was done exclusively through direct purchase mechanisms. In total, WFP procured 643 MT of millet in the first year, and 204 MT of millet and 48.7 MT of cowpeas in the second year through direct purchases.


Overall procurement times for sorghum averaged 205 days, seven of which were for contracting. Note that delivery times reflect the development goals of forward delivery contracts (direct purchases) from targeted smallholder suppliers. This was true for all such procurements using FDCs under this project. Overall procurement times for millet averaged 161 days, 38 of which elapsed during contracting. As noted above, delivery times reflect the development goals of forward delivery contracts (e.g., purchasing from targeted smallholder suppliers). Overall procurement times for cowpeas averaged 24 total days, with 13 spent in contracting and 12 spent in delivery.


Sorghum cost on average $270.81 per MT for commodities, plus $55.13 per MT for TSH. Millet cost $279.89 per MT for commodities, plus $53.54 per MT for TSH. Cowpeas cost $671.37 per MT for commodities, plus $52.60 per MT for TSH.


this section have been met WFP reporting was found to be in compliance with requirements. However, no national level historical supply (production) information was found for millet or sorghum, rather selected markets were used instead. WFP did not report any information required to assess market impact on cowpeas. WFP also did not report on historical market demand nor comment in depth on the impacts on beneficiaries and low-income consumers.



WFP locally sourced commodities directly from FOs. WFP set a maximum acceptable level of impurities that would meet WFP quality standards, for which it paid a price premium. This made the project attractive to farmers while encouraging them to appreciate the value of quality improvements.

39 "U.S. International Food Assistance Report 2010." USAID Food For Peace. USAID, n.d. Web. 9 Apr. 2012. <http://www.usaid.gov/our_work/humanitarian_assistance/ffp/annrep.html>. 40 U.S. Food Aid Tables FY2011, < http://www.fas.usda.gov/excredits/FoodAid/Reports/reports.html>.


Reporting Requirement Notes WFP further reported that for less-structured FOs the biggest gain was a realization of significantly increased incomes for members. WFP broke down impacts on FOs based on the FOs’ level of organization and coordination. Less organized FOs mostly saw only an increase in member incomes whereas WFP reported that well-organized FOs reaped many additional benefits including the ability to:

• increase the funds available to supply inputs (seeds, fertilizer, etc.) to their members;

• purchase grain from their members to augment their central stock; • finance part of the operating costs of the FO (electricity, water, staff

salaries, etc.); • purchase post-harvest equipment such as threshers and sifters, to

improve produce quality; and • provide some pre-financing to allow them to deal with household

expenses for which they might otherwise be forced to sell their produce prematurely (when prices were low).

WFP also received reports from farmers of: • increased smallholder farmer production to the guaranteed WFP

market, access to credit and banking, and the potential or opportunity to market to other buyers;

• improved awareness of the need for higher quality produce to earn a higher price on the market;

• enhanced capacity to plan how much land to cultivate to meet their commitments under the FDCs;

• increased sales and incomes of smallholder farmers; and • desire among other smallholder farmers to join the FO to benefit

from capacity building, aggregating their stocks, and producing higher quality commodities.

bb Markets

Market price data on millet and sorghum (collected before and after each procurement) support the conclusion that there is no market impact on millet and sorghum prices, since variations in those prices were negligible and the mean procurement size was small. Regarding cowpeas, time series data for the price of cowpeas were not provided in the WFP report. The team found that, while any impacts on the prices of cowpeas from the LRP project purchases are extremely unlikely, some impact on local market prices of cowpeas cannot be ruled out. WFP and the evaluation team concur in that no impact was found on market prices due to P4P procurement.

cc Low-income consumers Available data show relatively stable consumer prices for low-income consumers. However, data on cowpea prices were not available.

dd Program recipients WFP reported that USDA -funded LRP commodities supported nutritional rehabilitation. WFP delivered food to community health centers where patients received nutritional rehabilitation support. Semi-structured interviews and focus group discussions with local health center representatives and patients revealed that malnourished pregnant women, lactating women, and adults receiving treatment for HIV/AIDS and tuberculosis, all benefitted from the support of the project. Institutions and patients were appreciative and reported good overall recovery rates when the


Reporting Requirement Notes assistance was consistent. Patients additionally reported that the quantities received were sufficient for their needs. One patient receiving food assistance through the HIV/AIDS-TB program reported that after five months her weight increased by more than 15 kilograms.

III Time of delivery Overall procurement times for sorghum averaged 205 days, seven of which were for contracting. Note that delivery times reflect the development goals of forward delivery contracts (direct purchases) from targeted smallholder suppliers. This was true for all such procurements using FDCs under this project. Overall procurement times for millet averaged 161 days, 38 of which elapsed during contracting. As noted above, delivery times reflect the development goals of forward delivery contracts (e.g., purchasing from targeted smallholder suppliers). Overall procurement times for cowpeas averaged 24 total days, with 13 spent in contracting and 12 spent in delivery.

iii. Compare: IV Quality and safety assurances The evaluation team reviewed a sample of food quality and safety inspection

reports for millet, sorghum and cowpeas from the project. No food quality or safety issues were identified. Inspection of all commodities was carried out by Office des Produits Agricoles du Mali (OPAM), Mali’s cereals marketing agency.

FIGURE 1: MILLET PRODUCTION IN BANKASS

Source: OMA - Market Data, in Bishop, Helen and Pierre Traoré. Report on USDA-Funded Purchases 2009 – 2011, WFP Mali Final LRP Project Report Note: FCFA = Central African Franc


FIGURE 2: SORGHUM PRODUCTION IN KOUTIALA

Source: OMA - Market Data, in Bishop, Helen and Pierre Traoré. Report on USDA-Funded Purchases 2009 – 2011, WFP Mali Final LRP Project Report Note: FCFA = Central African Franc

FIGURE 3: WHOLESALE CEREAL PRICES IN BAMAKO, 2006 – 2011

Source: FAOSTAT. http://faostat.fao.org/

0

10

20

Jan-06 Jan-07 Jan-08 Jan-09 Jan-10 Jan-11

thou

sand

s of 2

000

CFA

fran

cs p

er 1

00kg

sa

ck

sorghum millet


FIGURE 4: DEFLATED MILLET PRICES IN BAMAKO 2006 – 2010


0

5

10

15

20


thou

sand

s of 2

000

CFA

fran

cs

per

100

-kg

sack

2006 2007

2008 2009

2010 MEAN


MOZAMBIQUE

WFP LOCAL AND REGIONAL PROCUREMENT PROJECT, EMERGENCY

WFP implemented an emergency LRP project in Mozambique, to relieve victims of floods and drought between April and July 2011. WFP’s country office in Mozambique issued regional tenders for the commodities, receiving successful bids from companies in Malawi, Mozambique and South Africa. Three procurements were regional, and one Mozambican company won the bid for the fourth procurement (this project thus has both local and regional procurement components). The regional markets were Lilongwé and Blantyre in Malawi (maize) and Durban in South Africa (cowpeas). The national market was Nampula for cowpeas. All commodities were delivered to the WFP warehouse in Beira. The project directly distributed maize and cowpeas to 132,000 beneficiaries in four provinces, Manica, Sofala, Tete and Zambézia.



WFP’s report shows that overall Malawi maize prices have historically fallen between March and June. During the procurement period in 2011, maize prices increased from January to March and decreased from March to June, following seasonal norms. September 2010 to August 2011 prices for maize in Malawi were lower than those of the previous three years, as a recent input subsidies program in Malawi increased maize production and consequently decreased maize prices overall. Annual Mozambican production of cowpeas normally averages 10 – 12,000 MT but poor rains brought down the quantities produced during the 2011 harvest to 6,000 – 7,000 MT. Prices of cowpeas in the North of Mozambique were 17 – 33 percent higher than normal a result41. In South Africa, cowpea prices averaged $376.4/MT between 2005 and 2009.


WFP purchased maize and cowpeas from markets in Malawi, South Africa, and Mozambique using competitive hard tendering, inviting WFP’s list of eligible suppliers for agricultural commodities to bid. Three competitive tenders were held. One tender for maize was issued on February 6, 2011, another for cowpeas was issued on March 3, 2011 and the last tender for maize was issued on April 18, 2011. The closing dates for the three tenders were February 22, 2011, March 9, 2011, and May 5, 2011, respectively. The key criteria used to select the winning suppliers were the cost of the commodity and the length of time that each supplier needed to deliver the commodity to the agreed-upon destination. Costs were also evaluated relative to import parity prices. Seven bids were received for each tender.


The evaluation team’s analyses of existing and available data indicate that this project had no likely impact on the prices of the affected commodities. Please see Annex 3 and the market impact chapter of the main body of the report for further detail. Conditions in the maize market during the LRP project included: aggressive buying by a range of suppliers; a lack of export ban imposed by the Government of Malawi unlike in previous years; institution of a new price floor by the Government of Malawi; and a sale of 30,000 MT of older maize stock from the

41 Data collected by the evaluation team in the field


Reporting Requirement Notes National Strategic Reserve Agency (NSRA), with the last two factors pushing prices down. At the same time, Malawian farmers held almost no stocks by this time of the year (several months after harvest), so the price effects were felt almost entirely among the large traders who were trading stock. These machinations likely had little effect on retail prices. A slight increase in the 2011 maize prices from March to April may be associated with the WFP procurement, which might have increased demand, at least among traders. Prices fell again till June. Comparing the tender prices and the prices of maize during the procurement period, it can be observed that tender prices were higher than the average market prices, supporting the conclusion that the uncommon market activities noted above were affecting prices for traders (but without data on retail price movements). The results overall show that local market prices in Malawi and Mozambique were rising but not likely due to the smaller LRP project procurements. Distribution area data show little or no effect from the LRP project. Annual Mozambican production of cowpeas averages 10 – 12,000 MT but poor rains reduced the 2011 harvest to 6,000 – 7,000 MT. As cowpeas are a cash crop, almost all of this production is marketed, entirely within Mozambique. All of the 136 MT of cowpeas that Maviga, a large Mozambican trader based in Nampula Province in northern Mozambique, supplied as LRP to WFP in Mozambique came from Nampula. Maviga had 46 MT of cowpeas in reserve, necessitating a purchase of 90 MT, which made up 1.5 percent of the 2011 marketed surplus of 6,000 MT. Prices for cowpeas during the procurement period did rise somewhat, but this may be due to the limited harvest that year. The supplier interviewed by the evaluation team argued that the company sells more or less the same quantity of cowpeas each year, which they are sure to sell sooner or later, and that the WFP contract did not have any immediate impact on supplies or prices.


The Government of Mozambique, a WTO member, takes an open trade stance with many of its policies and regulations, with few import duties on foodstuffs and other agricultural products. Informal cross-border trade is common between Malawi and northern Mozambique. Shipments of maize from Malawi that are destined for Zimbabwe are sometimes seized by monitors and kept as imports to Mozambique. In FY 2010, Mozambique benefited from U.S. international food assistance including 16,740 MT of cornmeal provided through the McGovern-Dole International Food for Education and Child Nutrition program; and 60,000 MT of wheat under a CCC-funded Food for Progress grant.42 In FY 2011, USAID programmed shipments of 36,900 MT of wheat through Title II and USDA programmed 5,600 MT of soy fortified cornmeal through the McGovern-Dole International Food for Education and Child Nutrition program to be sent to Mozambique.43 In Malawi, tariffs on maize grain were eliminated and import licenses are required to engage in maize trade. The government has no export ban in place but some reports indicate the Government of Malawi has stopped issuing export licenses in order to limit maize export. This information is consistent with the findings of

42 "U.S. International Food Assistance Report 2010." USAID Food For Peace. USAID, n.d. Web. 9 Apr. 2012. <http://www.usaid.gov/our_work/humanitarian_assistance/ffp/annrep.html>. 43 U.S. Food Aid Tables FY2011, < http://www.fas.usda.gov/excredits/FoodAid/Reports/reports.html>.


Reporting Requirement Notes the field survey where respondents reported that the government has in the past attempted to control maize exports and to issue export bans when domestic maize prices increase. U.S. food assistance activities in Malawi in FY 2010 included 6,690 MT of CSB provided through the USDA-funded McGovern-Dole International Food for Education and Child Nutrition program, and a Title II development program that provided 17,190 MT of CSB, vegetable oil, wheat and beans.44 In FY2011, USAID programmed to ship 21,200 MT of commodities under Title II programming and USDA programmed to ship 7,800 MT of CSB under the McGovern-Dole International Food for Education and Child Nutrition program.45


WFP Mozambique competitively procured 4,178.55 MT of maize regionally and 486 MT of cowpeas (including 350 MT regionally, and 136 MT locally).


Maize was procured in an average of 75 days, 14 for contracting and 61 for delivery. Cowpeas procurements averaged a total of 69 days, with 15 days for contracting and 54 for delivery.


Average costs for the hard tender procurements were as follows: • Maize costs averaged $197.12 per MT for commodities plus $61.39 per

MT for TSH and • Cowpea costs averaged $477.25 per MT for commodities plus $142.71

per MT for TSH.


this section have been met WFP reporting was found to be compliant with requirements.



In Malawi, because of the timing of the purchases (a few months after harvest, when farmers were less likely to hold stock) and the Government of Malawi’s release of 30,000 MT of maize stock into the market, the LRP project purchases were not likely to have affected producers. In South Africa, the size of the purchases were very small relative to production. Interviews with the large producers that supplied commodities indicated that the purchases had no impacts. Interviews with the Mozambican cowpea supplier revealed that the delivery of 136 MT of cowpeas to WFP did not have impacts on the local supply chain, since it did not change the quantity of cowpeas that the supplier usually purchases each year.

bb Markets The evaluation team’s analyses of existing and available data indicate that this project had no likely impact on the prices of the affected commodities in the procurement zones. Because of the timing of the purchases, the release of 30,000 MT of NSRA maize stock into the market in Malawi, the small quantities relative to market throughput, and the temporary nature of price changes, no impacts were judged likely from the LRP project purchases. Interviews with the

44 United States Agency for International Development and United States Department of Agriculture. U.S. International Food Assistance Report 2010. N.d. Web. 14 Apr. 2010. <http://www.usaid.gov/our_work/humanitarian_assistance/ffp/fy2010.ifarreport.pdf>. 45 U.S. Food Aid Tables FY2011, < http://www.fas.usda.gov/excredits/FoodAid/Reports/reports.html>.


Reporting Requirement Notes Mozambican cowpea supplier and South African supplier indicated no price impacts on local markets due to LRP. Please see Annex 3 and the market impact chapter of the main body of the report for further detail.

cc Low-income consumers WFP reported that there might have been a slight decrease in the price of cowpeas in one large market close to one of the distribution zones.46 The report suggests that lower retail prices may have benefited low-income consumers who purchased cowpeas. However, there is no evidence that the price decrease was caused by the LRP project. Other retail price data analyzed by the evaluation team show no price increases around the time of LRP distributions of commodities, indicating no other impact on low-income consumers.

dd Program recipients Respondents to a survey by WFP were asked to report the different types of beneficiaries (gender, orphan, elderly, adolescents, pregnant women, disabled, lactating women) covered by the program. WFP data show that the food distribution reached the targeted beneficiaries in their project zones, including women, adolescents, individuals with disabilities, and other disadvantaged groups. Some beneficiaries performed work projects in exchange for their food as well, providing community benefits such as rehabilitating roads, building public structures and fences, planting trees, and manufacturing bricks. In total, 132,000 individuals in four provinces in Mozambique (Manica, Sofala, Tete and Zambézia) benefitted from this project.

III Time of delivery Maize was procured in an average of 75 days, 14 of which elapsed in contracting and 61 in delivery. Cowpea procurements averaged a total of 69 days, with 15 days for contracting and 54 for delivery.

iii. Compare: Quality and safety assurances Intertek, an independent commodity inspector, conducted verification inspections

for the WFP project in Mozambique. The national laboratory Laboratorio Nacional de Higiene de Alimentos e Aguas (LNHAA), conducted quality testing as part of Intertek’s service, and the commodities were found to be appropriate for human consumption. Aflatoxin was not detected in the sampled commodities.

TABLE 1: HISTORIC MARKET SUPPLY AND DEMAND IN MALAWI (SOURCE OF REGIONAL PROCUREMENT)

CEREAL SUPPLY/DEMAND BALANCE FOR THE 2010/11 MARKETING YEAR (April/March)

Wheat Rice Coarse Grains Total Cereals

Cereal supply and utilization data In thousands of MT 2010/11 Domestic Availability 17 230 2,527 2,774 2010/11 Utilization 484 488 2,661 3,633 2010/11 Import Requirement 467 258 134 859 Source: Global Information and Early Warning Systems (GIEWS)

46 Mutarara market in Tete Province


FIGURE 1: HISTORICAL MARKET PRICE MOVEMENTS IN MALAWI

Source: USDA Local and Regional Procurement Pilot Project, Evaluation Report –Final LRP Project Report World Food Programme/Mozambique

FIGURE 2: SEASONAL INDEX PRICE OF MAIZE IN LILONGWE, MALAWI (PROCUREMENT AREA)

Source: USDA Local and Regional Procurement Pilot Project, Evaluation Report –Final LRP Project Report World Food Programme/Mozambique

Pric

e (K

wac

ha/K

g)

Maize

Liwonde 2011

Lilongwé Average (2007-2010)

0.00

20.00

40.00

60.00

80.00

100.00

120.00

140.00

160.00

Seasonal Index

White Maize


FIGURE 3: MONTHLY PRICES OF MAIZE IN MALAWI

Source: Adapted from WFP data file, Mozambique country office


NICARAGUA

FABRETTO CHILDREN’S FOUNDATION LOCAL PROCUREMENT PROJECT, DEVELOPMENT

Fabretto Children’s Foundation (FCF) is a non-profit organization working to provide nutritional support and educational opportunities to impoverished children in Nicaragua. The USDA-supported school lunch program provided school lunches that addressed malnutrition and encouraged school attendance.

The project targeted 20,933 beneficiaries and had a total cost of $677,363. The project purchased (in MT): butter (0.4); carrots (12.1); cheese (17.9); corn flour (43.0); cream (12.7); eggs (9.9); green peppers (5.2); pinolillo - a sweetened corn drink (7.1); onions (13.6); plantains (44.8); potatoes (33.0); red beans (1.7); bananas (40.3); sugar (16.1); tomatoes (25.2); and vegetable oil (14.5). The commodities were procured through direct purchases with FOs and through competitive tenders using the Nicaraguan agricultural commodities exchange (BAGSA). The suppliers included local vendors, producers, and an umbrella farmers’ organization.


i. For each market and commodity procured, describe:


Historical maize and bean production are shown in Figure 1. The historical price for red beans is shown in Figure 2. Data presented in Figure 2 suggests that red bean prices during the project period were abnormally high, as a result of weather and economic shocks, including Hurricane Agatha in June of 2010. Historical prices for wholesale white maize are also shown in Figure 2, with a relatively steady trend. Historic market price information was found for maize and beans. No historic price information was found for other commodities procured FCF-implemented LRP project. Vegetable oil production in Nicaragua was extremely limited and most vegetable oil products were imported. Palm oil production has varied over previous decades. Data is available for 2008 – 2009 for palm oil production. Though palm oil is exported (approximately 11,000 MT annually), approximately 35,000 MT is also imported annually to meet country needs. Palm oil sources vary, but the majority of imports is from Central American producers and re-packaged for sale in Nicaragua. FCF also reported broad production data for sugar, eggs, milk, potatoes, plantains, and bananas.


FCF procured fruits, vegetables and dairy products through direct purchases with two cooperatives, selected based on their ability to provide the commodities on specific dates and in the quantities required. The ASOGYPTBM cooperative, located in Quilali, Estelí department, provided the fruits and vegetables, while the 5 de Junio cooperative, located in Las Sabanas, Madriz department, provided the dairy products. The bulk of corn flour, cooking oil, and red beans was purchased through open tenders using BAGSA. For the purchases of each of these products, FCF notified the exchange of the quantity required and product specifications, including requirements that the suppliers submit their commodities for quality


Reporting Requirement Notes testing. BAGSA notified its member vendors, as well as the wider agricultural vendor community in Nicaragua, and sales were made via an auction format. The quantities that FCF purchased through the exchange, however, were very small compared with those of other commercial BAGSA users. Thus, contenders to sell to FCF were very limited and three times, only one vendor bid on the tender. This was particularly true for the sale of pinolillo and corn flour, which were purchased frequently and often in small quantities because of their limited shelf-life. FCF also procured small amounts of corn flour, cooking oil, red beans, sugar, vegetable oil and pinolillo from suppliers through direct purchases.


The evaluation team’s review of existing and available data indicate that the amounts obtained for each commodity were much less than one percent of national production figures. Further, FCF purchases were relatively small and consisted of numerous purchases over the course of a year. The evaluation team’s analyses of existing and available data indicate that this project had no likely impact on the affected commodities’ prices. Please see Annex 3 and the market impact chapter of the main body of the report for further details.


FCF reported that in 2011 the Ministry of Finance, Industry, and Commerce allowed duty-free importation of 20,000 MT of corn flour to stabilize prices. An effective export ban for red beans was lifted in 2011. In September 2011, the Government of Nicaragua lifted all import duties on beans as well. The evaluation team found that trade policies (import tariffs, import quotas, and export subsidies) of the Government of Nicaragua limited trade between the international and domestic markets, preventing market integration. A focus on exports prioritizes economic growth over stabilizing weak internal food security, while international donor projects tend not to be institutionalized.47 Nicaragua received U.S. government food assistance in 2010 through a Title I -funded Food for Progress grant that supplied 33,000 MT of soybean meal and yellow corn and through a CCC-funded Food for Progress grant that supplied 9,750 MT of yellow corn as well as 12,020 MT of commodities through the McGovern-Dole International Food for Education and Child Nutrition program.48


The project purchased (in MT): butter (0.4); carrots (12.1); cheese (17.9); corn flour (43.0); cream (12.7); eggs (9.9); green peppers (5.2); pinolillo - a sweetened corn drink (7.1); onions (13.6); plantains (44.8); potatoes (33.0); red beans (1.7); bananas (40.3); sugar (16.1); tomatoes (25.2); and vegetable oil (14.5). Many of these individual procurements were for products that offered no comparison (as no other project procured these types of commodities). Only in the case of cereals (corn flour), pulses (red beans), and vegetable oil are the procurements included in the comparison dataset for this report.

47Sahley, Caroline; Crosby, Benjamin; Nelson, David and Vanderslice, Lane. 2005. “The Governance Dimensions of Food Security in Nicaragua,” USAID study, downloaded at http://pdf.usaid.gov/pdf_docs/PNADE106.pdf 48United States Agency for International Development and United States Department of Agriculture. U.S. International Food Assistance Report 2010. N.d. Web. 14 Apr. 2010. <http://www.usaid.gov/our_work/humanitarian_assistance/ffp/fy2010.ifarreport.pdf>.




Average overall times to procure were as follows: • Corn flour took on average 10 days total, with 6 days for contracting

and 4 in delivery; • Red beans took on average 14 days total, with 6 in contracting and 8

in delivery; and • Vegetable oil took on average 8 days total, with 5 in contracting and

3 in delivery. Thirty additional, very small procurements were also made of vegetables, fruits, dairy, eggs, sugar and a drink mix (pinolillo); these procurements are not compared in the time and cost datasets of this report because no other projects procured similar commodities. However, these commodities had roughly the same contracting and delivery times as the bulk of procurements listed above.

VI Total cost of procurement including storage, handling, transportation, and administrative costs

Average commodity costs per MT for procurement were as follows: Hard tender: Corn flour cost $1,091.17 per MT; Red beans cost $1,297.95 per MT; Vegetable oil cost $2,062.41 per MT; Butter cost $7,058.33 per MT; Cream cost $2,997.92 per MT; Sugar cost $840.65 per MT; and Pinolillo cost $1,682.77 per MT. Direct Purchase: Corn flour cost $815.85 per MT; Red beans cost $1,301.68 per MT; Vegetable oil cost $1,568.50 per MT; Carrots cost $966.42 per MT; Cheese cost $3,258.58 per MT; Eggs cost $1,442.43 per MT; Green peppers cost $2,300.55 per MT; Onions cost $663.46 per MT; Pinolillo cost $2,442.71 per MT; Plantains cost $613.08 per MT; Potatoes cost $865.72 per MT; Ripe bananas cost $354.38 per MT; Sugar cost $633.31 per MT; and Tomatoes cost $1,111.83 per MT. All TSH costs were $220.00 per MT.

ii. Assess:


FCF’s reporting was found to be compliant with requirements. Historic market price information was found for maize and beans only.



The procurements made under this agreement caused small-scale increases in production by farmers aligned with the supplier organizations (in the direct purchase procurements), though these increases were negligible due to the very small quantities FCF purchased weekly. FCF worked with two cooperatives for the dairy, fruits and vegetables purchases, and provided them training and technical assistance over the course of the project. Through their


Reporting Requirement Notes participation, these organizations became better equipped to undertake procurement contracts for donor or other clients in the future, with greater ability to manage commodity purchases and logistics. FCF noted that the relatively small procurements tendered through the commodity exchange (hard tender procurements) received fewer bids than larger procurements would have received, indicating a lack of interest in these small procurements by the larger producers and traders. As a result, impacts on larger producers were likely very low, nor were they targeted for support as part of the project.

bb Markets

FCF reported that for fresh fruits and vegetables, as well as dairy products, production by the contracted FOs increased marginally to meet the extra demand from the LRP project. The local market prices of corn flour and red beans for each time period (the week before, the week of, and the week after procurement) show negligible or no variation. Red bean and corn flour prices are extremely stable over the period. This coherence and degree of stability leads to the evaluation team’s conclusion that no market impacts were likely as a result of corn flour and red bean purchases made under the project. Please see Annex 3 and the market impact chapter of the main body of the report for further details.

cc Low-income consumers The impact on low-income consumers would primarily be felt in any price increases in distribution markets. However, as noted above, prices of staple commodities were stable during the period of procurement. FCF’s purchases were also very small relative to national markets, likely producing no impact on low income consumers through price rises or volatility. As an example, FCF purchased 47 MT of beans through the BAGSA bidding system, in comparison to the 463 million MT produced by the entire country in 2010.

dd Program recipients FCF reported that the targeted beneficiaries, Nicaraguan school children, benefitted the most from this project, as they received up to 50 percent of their caloric intake during school days through USDA-funded LRP distributions. FCF noted a direct and positive correlation between the school lunch program and attendance; thus, children attended school and their dietary needs were at least partially addressed as a result of this program.

III Time of delivery Average overall times to procure were as follows: • Corn flour took on average 10 days total, with 6 days for contracting

and 4 in delivery; • Red beans took on average 14 days total, with 6 in contracting and 8

in delivery; and • Vegetable oil took on average 8 days total, with 5 in contracting and

3 in delivery. Thirty additional, very small procurements were also made of vegetables, fruits, dairy, eggs, sugar and a drink mix (pinolillo); these procurements are not compared in the time and cost datasets of this report (because no other projects procured similar commodities) but had roughly the same contracting and delivery times as the bulk of procurements listed above.



iii. Compare:

IV Quality and safety assurances Commodity quality testing was carried out by Laboratorios Bengoechea and Labtec (Laboratorio Tecnico), including testing on cooking times (for freshness), aflatoxin, chemical content, and microbiological residuals. FCF submitted inspection reports from the testing labs, showing no issues with food quality or safety. Non-perishables purchased on the BAGSA exchange needed to meet minimum standards established by the Ministry of Health in regard to permitted levels of aflatoxin. Food handlers (usually volunteer mothers in most schools, as well as paid cooks in centers) received training on food handling and safety, particularly in food storage and how to identify freshness at the time of purchase.

FIGURE 1: BEANS AND MAIZE PRODUCTION IN NICARAGUA, 2003-2011

Source: Dirección de Estadísticas, Ministerio Agropecuario y Forestal (MAG-FOR), Nicaragua

0

2,000

4,000

6,000

8,000

10,000

12,000

14,000

Thou

sand

s Qui

ntal

s

Beans MaizeDireccion de Estadisticas MAG-FOR


FIGURE 2: RED BEANS AND WHITE MAIZE PRICES (WHOLESALE), 2008-2011, MANAGUA, NICARAGUA

Source: FAO GIEWS Food Price Data and Analysis Tool at http://www.fao.org/giews/pricetool2/


NIGER

CRS LOCAL PROCUREMENT PROJECT, EMERGENCY

In response to the food insecurity situation of 2011, CRS implemented Project VOICE in Niger with the support of USDA. The CRS LRP project operated in an area that was adjacent to the region in which the USDA-funded Mercy Corps project was implemented. The CRS emergency food security voucher program was carried out from January through September 2011. Through Project VOICE, CRS supported a population of 140,758 drought-affected people in 111 villages in the Tillaberi and Zinder regions. The program targeted 21,000 of the most vulnerable households with monthly food voucher distributions. The suppliers were 38 local vendors. Through the project, beneficiaries purchased 6,137 MT of millet, sorghum, beans, vegetable oil, rice and maize.

Reporting Requirement Notes i. For each market and commodity procured, describe: I Prevailing and historic

supply, demand, and price movements of the market

Project VOICE conducted interventions in three departments in the Tillaberi and Zinder regions, namely Ouallam, Mirriah, and Gouré. Rural primary production forms the largest part of Niger’s economy, accounting for about 45 percent of GDP and supporting about 80 percent of the population.49 Rainfed subsistence agriculture and extensive livestock rearing comprise almost all this economic acti-vity.50 Farmers mostly produce cereals and other crops to meet their household’s nutritional needs, with millet accounting for 70 percent of staple-crop production, sorghum for 16 percent and cowpeas for 13 percent. However, Niger’s substantial production of millet never meets its consumption needs: the country imports it as a matter of course, mostly from Nigeria. The result of a fast growing population on degrading agricultural soils with lower and more erratic rainfall is recurring famine. At the national level, Niger produced nearly 3 million MT of coarse grains in 2010, and 1.8 million MT of cowpeas.51 The Food Balance Sheet for cereals in Niger is shown in Table 1, below, showing a net deficit. The deficit is more severe in some regions than others, including the regions targeted in this LRP project, as the drought was more severe in those regions. The market is well-integrated for many commodities and prices from markets in target areas are mostly parallel with those in Niamey. The evaluation team found that Niger’s cereal prices typically followed an annual cycle, with prices at their lowest during harvest and just afterwards (Figure 1). Thereafter, prices rose as scarcity increased until around August, when firm indications of the size of the next harvest began to emerge. Variations in harvest levels make seasonal price profiles unpredictable. The prices of millet, sorghum, and maize – close dietary substitutes – mostly rose and fell together in the competitive Niger market (Figure 2). All of these commodities were available on the market from domestic sources or were readily imported, mostly from Nigeria. CRS reported price data in Niamey and Ouallam in the Tillaberi region and Zinder and Gouré in the Zinder region. These data are discussed in this section, with the addition of secondary source data on the markets.

49 http://siteresources.worldbank.org/IDA/Resources/73153-1285271432420/IDA_AT_WORK_Niger_2010.pdf 50 Niger has 270,000 hectares of irrigable land, of which just over half lie in the valley of the river Niger. 51 FAO GIEWS Food Price Data and Analysis Tool http://www.fao.org/giews/pricetool2/


Reporting Requirement Notes The price trends for maize, millet and sorghum in the Niamey market for the last three years showed very similar movements over the same periods; increases in July and August were seen each year. While voucher distributions did not take place in communities that rely on the Niamey market, this market is important because it is the hub that supplies markets in the Ouallam Department. It can also be utilized as a reference (test market) for the functioning of a “normal” market (i.e., one that did not receive a voucher intervention). The prices in the Ouallam market saw nearly the same trends as those in Niamey—stability from March to July 2010 and 2009, and lower prices in September. This market supplied all of the small markets within departments where voucher swaps took place. Higher prices in Ouallam, when compared with Niamey, are linked to transportation costs (i.e., the additional cost of transporting commodities from Niamey to Ouallam, an increase in gasoline prices, and the arrival of the rainy season that made transport to local markets more difficult). The Zinder and Tillaberi markets were less stable during the LRP project time period for all three cereals; an example of this is shown in Figure 3 for millet. Part of this instability was linked to the 2011 drought and resulting production shortfalls in the zones that supply the commodities that beneficiaries in Zinder and Tillaberi generally purchase. Because of this, the price situation in 2011 cannot be linked directly to the project intervention.


CRS used several criteria to select its vendors. Any trader wishing to participate in the program had to be legally recognized under the laws of the country through the trade union, have a bank account in order to facilitate reimbursement, and to the extent possible, have a presence in the departments of the project intervention zone. Taking these criteria into account, all 38 vendors who applied qualified to participate in the program. CRS signed contracts with vendors that had clauses limiting speculation and the selling of products of questionable quality. Voucher recipients could then choose the vendor from among those in their areas.


CRS reported that their price analysis does not reflect the impact on producer prices because producers in Niger sell commodities in the months immediately after the harvest (October) and not during the period of the project intervention. CRS further reported that, given the information collected for market prices over the last three years, the intervention did not appear to have a significant impact on commodity prices. Additionally, as outlined below, consumers seemed to benefit in ways other than price, such as beneficiaries’ autonomy to choose which commodities to purchase, and the quality and consistency of sufficient supply. Given three-year market price data analyzed for the evaluation, the project does not appear to have had a significant impact on commodity prices. And as outlined below, consumers seemed to benefit in ways other than price, such as the beneficiaries’ autonomy to choose which commodities to purchase, quality, and consistent, sufficient supply. Please see Annex 3 and the market impact chapter of the main body of the report for further details.

III Each government market interference or other activity of the donor country that might have significantly affected the supply or demand of the eligible commodity in the area at

The evaluation team found that the Government of Niger largely allows the market to allocate cereals and cowpeas. Food flows freely between markets and across borders with little hindrance, though sometimes the Burkinabè government imposes bans on the export of cereals to Niger and other countries. CRS reported that although there were food programs being conducted by the Government of Niger, its close coordination with the Office of the Prime


Reporting Requirement Notes which the local or regional procurement occurred

Minister, Food Crisis Section helped to ensure that these distributions did not occur in intervention zones where CRS was working and had no impact on the supply or demand of the commodities purchased through the voucher program. CRS has received U.S. Government funding for several food assistance programs in Niger. However, CRS reported that, due to coordination with its staff working on the USAID-funded Multi-Year Assistance Program (MYAP) and Single-Year Assistance Program (SYAP), the MYAP and SYAP interventions, including blanket feeding, literacy feeding, and nutritional rehabilitation, did not affect commodity supply or demand in the LRP implementation zones. In 2010, Niger received over 30,000 MT of cereals, pulses, oils, and CSB under a Title II emergency program and an additional 15,000 MT were provided under a Title II development program.52 Due to continuing food insecurity, 2011 programmed shipments from USAID included over 42,000 MT for Niger as well.53


At the end of four rounds of voucher distributions and redemptions, the following products were purchased by beneficiaries: Cereals: 4,760.52 MT; Cowpeas: 792.88 MT; and Vegetable oils: 583.54 MT.


The average total procurement time for all three commodities through vouchers was 38 days.


Average commodity costs were as follows: • Cereals cost $503.56 per MT; • Cowpeas cost $465.52 per MT; and • Vegetable oil cost $2,579.00 per MT.

All average TSH costs were $13.40 per MT.


this section have been met CRS reporting was found to be compliant with requirements.



CRS reported that evaluating the impact of distributions on producers is difficult for two reasons: Most of the commodities purchased were produced outside of Niger with the exception of cowpeas and some of the millet and sorghum. Local production of millet and sorghum was weak and made up only a small quantity of the purchased cereals, according to CRS’ final report; beneficiaries chose other commodities instead, and these originated outside of Niger. Producers are generally small-scale farmers who sell their grain in order to satisfy the needs of their households, typically just after harvest in October (outside of the project’s purchasing timeframe). For these reasons, no impacts on producers were found.

bb Markets The evaluation team’s analyses of existing and available data indicated that this project had no likely impact on the price of affected commodities. Please see

52 United States Agency for International Development and United States Department of Agriculture. U.S. International Food Assistance Report 2010. N.d. Web. 14 Apr. 2010. <http://www.usaid.gov/our_work/humanitarian_assistance/ffp/fy2010.ifarreport.pdf>. 53U.S. Food Aid Tables FY2011, < http://www.fas.usda.gov/excredits/FoodAid/Reports/reports.html>.



Annex 3 and the market impact chapter of the main body of the report for further details.


CRS identified the following possible impacts on consumers who did not receive vouchers:

• Non-beneficiaries benefited from the intervention because of established patterns of food/grain-sharing within communities;

• The increased attendance at markets on the days of voucher distributions favor non-beneficiaries by giving them bargaining power with vendors who are not part of the program; and

• Low-income consumers can be particularly impacted in the marketplace if prices rise or fall. However, given the evaluation team’s analysis of no likely impact on the price of affected commodities, impacts on low-income consumers were likely negligible or non-existent.

dd Program recipients CRS reported that during the program some vendors hired additional employees, creating economic opportunities for others in the community. Vendors without bank accounts opened accounts and now use them as a method of saving money. The LRP project allowed beneficiaries to receive food that was culturally appropriate. Beneficiaries preferred vouchers to direct distributions because they felt that foods directly distributed through other programs were not the foods to which they were accustomed. As further reported by CRS, the impacts on targeted beneficiaries include:

• Improved food availability within the household during the period; • The return of family members who had sought work outside of the

community, or an incentive to stay within the community; • Improved child nutrition; and • Strengthened community solidarity, which included ration-sharing. For

example, after one voucher distribution, a family could buy about 70 kg of food and give away 5 kg.

III Time of delivery The average total procurement time for all three commodities through vouchers distributions was 38 days.

iii. Compare: IV Quality and safety assurances CRS submitted inspection reports and a sample of these was reviewed by the

evaluation team. The evaluation team found no food quality or safety issues mentioned in the reports. The government laboratory LANSPEX was contracted to conduct bimonthly testing on millet, sorghum, beans, and maize. The laboratory tested for humidity levels, the presence of insects, foreign components, broken grains, and the presence of other types of grain, mold, and aflatoxin.


TABLE 1: NIGER’S CEREALS BALANCE SHEET, 2009-2010, IN THOUSANDS OF TONS

Rice Wheat

Coarse cereals Total

Net production 59 1 2,910 2,968 Net consumption 263 88 3,028 3,379 Net surplus -204 -87 -120 -411

Source: Evaluation de la campagne agricole 2009/2010 Note: “Coarse cereals” = millet, sorghum, maize and fonio.

FIGURE 1: DEFLATED NIAMEY MILLET PRICES 2002-2010, 1996 CFA FRANCS PER KG

Sources: SIMA (prices) and l’Institut National de la Statistique (CPI deflator)

100

150

200

250

300

Aug

Sep

Oct

Nov De

c

Jan

Feb

Mar Ap

r

May Jun Jul

2002-03

2003-04

2004-05

2005-06

2006-07

2007-08

2008-09

2009-10

MEAN


FIGURE 2: RETAIL CEREAL PRICES IN NIAMEY 2009-2011, USD PER KG


FIGURE 3: RETAIL MILLET PRICES IN THREE MARKETS, 2009-2011, USD PER KG



NIGER

MERCY CORPS LOCAL PROCUREMENT PROJECT, EMERGENCY

Mercy Corps implemented the Localized Immediate Drought Response (LIDR) project, a USDA-financed LRP project in parts of the Filingué Department of Tillabéri Region of Niger. Mercy Corps’ food aid distributions were adjacent to but did not overlap geographically with the USDA-funded LRP project carried out by CRS (above). Mercy Corps organized the distribution of 4,699 MT of maize and 472 MT of cowpeas as well as millet, cooking oil, and salt to 78,608 beneficiaries. Each beneficiary household received rations calculated to meet 80 percent of its monthly needs for five months. The first month of the program was at the end of the 2010 lean season (September 2010) and the program resumed for four months from April to July 2011. Competitive procurement of maize, cowpeas, and millet for these distributions took place in September 2010, March 2011, and July 2011, with about three-quarters of the total procured in March 2011. Mercy Corps allocated the oil and salt through a voucher system; the vouchers were for a fixed volume of oil and a fixed weight of salt, rather than denominated in currency.

Reporting Requirement Notes i. For each market and commodity procured, describe:

I Prevailing and historic supply, demand, and price movements of the market

Rural primary production forms the largest part of Niger’s economy, accounting for about 45 percent of GDP and supporting about 80 percent of the population.54 Rainfed subsistence agriculture and extensive livestock rearing comprise almost all this economic activity.55 Farmers mostly produce cereals and other crops to meet their household’s nutritional needs, with millet accounting for 70 percent of staple-crop production, sorghum for 16 percent and cowpeas for 13 percent. However, Niger’s substantial production of millet never meets its consumption needs: the country imports it as a matter of course, mostly from Nigeria. The result of a fast growing population on degrading agricultural soils with lower and more erratic rainfall is recurring famine. At the national level, Niger produced nearly 3 million MT of coarse grains in 2010, and 1.8 million MT of cowpeas.56 The Food Balance Sheet for cereals in Niger is shown in Table 1, below, showing a net deficit. The deficit is more severe in some regions than others, including the regions targeted in this LRP project, as the drought was more severe in those regions. The market is well-integrated for many commodities and prices from markets in target areas are mostly parallel with those in Niamey. Niger’s cereal prices typically follow an annual cycle, with prices at their lowest during the harvest and just afterward (Figure 1). Thereafter, prices rise as scarcity increases until around August, when firm indications of the size of the next harvest emerge. In successive years, however, relative harvest levels in different parts of the country and in Nigeria make seasonal price profiles unpredictable. The prices of millet, sorghum, and maize––close dietary substitutes––mostly rise and fall together in the competitive Niger market (Figure 2). All of them are available on the market from domestic sources or are readily imported, mostly from Nigeria.

54 http://siteresources.worldbank.org/IDA/Resources/73153-1285271432420/IDA_AT_WORK_Niger_2010.pdf 55 Niger has 270,000 hectares of irrigable land, of which just over half lie in the valley of the river Niger. 56 FAO GIEWS Food Price Data and Analysis Tool at http://www.fao.org/giews/pricetool2/


Reporting Requirement Notes Cowpeas and millet are produced in the area, but in Filingue department in a normal year, production is less than a third of the area’s consumption needs. Much of the production from the region is also exported after harvest, when farmers are seeking immediate cash for their goods. Surpluses from the south of Niger can mitigate food insecurity within the country during good harvests, but harvests in 2009, 2010 and 2011 were low due to drought.


Thirteen procurements were made by competitive tender. Eleven procurements received nine or more bids. Two procurements received three or fewer bids (Table 2).


Barrett et al. (2011: 4-5)57 found that the procurement of millet, which they attribute not only to LRP financed by USDA but also to WFP procurement taking place at the same time, caused statistically insignificant changes in the price of millet since both the USDA-funded and WFP procurements took place at the same time. They noted that the LRP procurement of cowpeas did not show a statistically significant effect. However, the authors found that Mercy Corps’ local purchases of maize appear to have immediately increased maize prices by 6.7 percent, increasing to 11.6 percent after one month and falling to 8.8 percent after two months. This fluctuation appears plausible as the maize contracts accounted for about 90 percent of total procurement. Most of the maize was purchased in Niger, but produced in Benin. Mercy Corps did not conduct market research in Benin (or in any source countries outside Niger) to establish market conditions. Barrett et al. (2011: 5) found a small statistically significant increase in the volatility of cowpea and maize prices attributable to Mercy Corps’ distributions, but none attributable to procurement (either on the part of Mercy Corps alone for maize or cowpeas, or due to procurements made by Mercy Corps and WFP for millet). From available price data, the evaluation team determined that the purchase of millet in the local market did not appear to negatively impact producers or consumers. This can be explained by the fact that it was purchased in a small quantity and was used to restock community cereal banks rather than for direct distribution. For the other commodities purchased under the LRP project, some possibility of small price impacts was found. The evaluation team’s analyses of existing and available data indicate that this project had a possible impact on the prices of affected commodities. Please see Annex 3 and the market impact chapter of the main body of the report for further details.


The Government of Niger largely allows the market to allocate cereals and cowpeas. Food flows freely between markets and across borders, though sometimes the Burkinabè government imposes bans on the export of cereals. There is a government safety net program or early warning program, but it does not include Mercy Corps’ project site of Filingue. Africare had a USAID-funded MYAP but in different communes than those of the USDA LRP project. In general, the larger food security programs were implemented to the east of the LRP project site. Mercy

57The Estimated Price Impacts, Timeliness, and Cost-Effectiveness of the Local and Regional Procurement (LRP) of Food Aid, Funded by USDA, Mercy Corps, Niger


Reporting Requirement Notes Corps chose to implement in Filingue because of recurring drought in recent years, resulting in insufficient harvests that have left households highly food insecure. In 2010, Niger received over 32,000 MT of cereals, pulses, oils, and FBFs under a Title II emergency program and an additional 15,000 MT were provided under a Title II development program.58 Due to continuing food insecurity, 2011 programming from USAID included over 42,000 MT for Niger as well.59


Quantity and type of competitive and limited competitive tender procurements are shown in Table 2 below.


The average number of days for 13 procurements were as follows: • Cowpeas took an average of 61 days total, with 56 for contracting

and five for delivery; • Maize took an average of 61 days total, with 56 for contracting and

five for delivery; • Millet took an average of 44 days total, with 38 for contracting and

six for delivery; and • Salt and vegetable oil took an average of 49 days each for the

voucher process.


Average commodity costs per MT for the procurements were as follows: Hard tenders: Maize cost $386.23 per MT plus $109.47 per MT for TSH; Cowpeas cost $543.55 per MT plus $121.18 per MT for TSH; and Millet cost $419.96 per MT plus $0 per MT for TSH. Vouchers: Vegetable oil cost $2,610.60 per MT and Salt cost $508.80 per MT. All TSH costs for voucher procurements were $20.61 per MT.

ii. Assess:


Mercy Corps’ reporting was found to be compliant with requirements.



While this program did not directly target local and regional producers or measure the impact of the program on this group, the main impact of the procurement methodology was an increase in demand for their products and increased sales.

bb Markets Barrett et al. (2011: 4-5)60 found that the procurement of millet, which they attribute not only to LRP financed by USDA but also to WFP procurement taking place at the same time, caused statistically insignificant changes in the price of millet since both the USDA-funded and WFP procurements took place at the same time. They noted that the LRP procurement of cowpeas did not show a statistically significant effect. However, the authors found

58United States Agency for International Development and United States Department of Agriculture. U.S. International Food Assistance Report 2010. N.d. Web. 14 Apr. 2010. <http://www.usaid.gov/our_work/humanitarian_assistance/ffp/fy2010.ifarreport.pdf>. 59U.S. Food Aid Tables FY 2011< http://www.fas.usda.gov/excredits/FoodAid/Reports/reports.html>. 60The Estimated Price Impacts, Timeliness, and Cost-Effectiveness of the Local and Regional Procurement (LRP) of Food Aid, Funded by USDA, Mercy Corps, Niger


Reporting Requirement Notes that Mercy Corps’ local purchases of maize appear to have immediately increased maize prices by 6.7 percent, increasing to 11.6 percent after one month and falling to 8.8 percent after two months. This fluctuation appears plausible as the maize contracts accounted for about 90 percent of total procurement. Most of the maize was purchased in Niger, but produced in Benin. Mercy Corps did not conduct market research in Benin (or in any source countries outside Niger) to establish market conditions. Barrett et al. (2011: 5) found a small statistically significant increase in the volatility of cowpea and maize prices attributable to Mercy Corps’ distributions, but none attributable to procurement (either on the part of Mercy Corps alone for maize or cowpeas, or due to procurements made by Mercy Corps and WFP for millet). From available price data, the evaluation team determined that the purchase of millet in the local market did not appear to negatively impact producers or consumers. This can be explained by the fact that it was purchased in a small quantity and was used to restock community cereal banks rather than for direct distribution. For the other commodities purchased under the LRP project, some possibility of small price impacts was found. The evaluation team’s analyses of existing and available data indicate that this project had a possible impact on the prices of affected commodities. Please see Annex 3 and the market impact chapter of the main body of the report for further details.

cc Low-income consumers Vendors approved for the voucher program stated that because the distributions were time-bound and due to the fact that only a segment of the population benefited from them, the distributions largely did not affect business at their stalls or shops, though they did notice a slight decrease in their customer base. Possible market price impacts were found to be statistically insignificant in the Cornell study. Had there been larger price impacts, low-income consumers could have been negatively affected, but no evidence was found to suggest that such larger price impacts occurred.

dd Program recipients Over 72,000 direct beneficiaries received rations, repeatedly throughout the project period. Roughly 50 percent of beneficiaries were women, and approximately 58 percent were under 18 years of age. Ration sizes were designed to meet 80 percent of the energy needs of chronically food-insecure households for each month when food was delivered. Beneficiaries reported that one effect of the food distribution was that they were able to avoid the coping mechanisms of migration, working on others’ farms for day wages, or selling household assets to purchase food.

III Time of delivery The average number of days for 13 procurements were as follows: • Cowpeas took an average of 61 days total, with 56 for contracting

and five for delivery; • Maize took an average of 61 days total, with 56 for contracting and

five for delivery; • Millet took an average of 44 days total, with 38 for contracting and

six for delivery;


Reporting Requirement Notes • Salt took an average of 49 days total, with 43 for contracting and

six for delivery; and • Vegetable oil took an average of 49 days total, with 43 for

contracting and six for delivery.

iii. Compare: IV Quality and safety assurances To test commodities, Mercy Corps Niger issued a tender to select an

inspection company to perform commodity testing before distributions took place. Commodities that were purchased directly by Mercy Corps were tested in Mercy Corps' temporary warehouses. The inspection company that was selected came to the warehouse as commodities were delivered and took one sample for each 100 MT. For the voucher component, samples were taken from selected vendors' warehouses. Testing was conducted over five to seven days at three levels with three different laboratories (one regional and two national). Tests included: aflatoxin/mycotoxin, impurities, chlorine testing, and the acid rate, as relevant for each commodity. After testing, a report was provided to Mercy Corps. No major problems were found and distributions proceeded as planned. When necessary, fumigation was conducted during purchasing and distribution periods.

TABLE 1: NIGER’S CEREALS BALANCE SHEET, 2009-2010, IN THOUSANDS OF TONS

Rice Wheat

Coarse cereals Total

Net production 59 1 2,910 2,968 Net consumption 263 88 3,028 3,379 Net surplus -204 -87 -120 -411

Source: Evaluation de la campagne agricole 2009/2010 Note: “Coarse cereals” = millet, sorghum, maize and fonio.


FIGURE 1: DEFLATED NIAMEY MILLET PRICES, 2002-2010, IN 1996 CFA FRANCS PER KG

Sources: SIMA (prices) & l’Institut National de la Statistique (CPI deflator)

FIGURE 2: RETAIL CEREAL PRICES IN NIAMEY 2009-2011, USD PER KG


100

150

200

250

300

Aug

Sep

Oct

Nov De

c

Jan

Feb

Mar Ap

r

May Jun Jul

2002-03

2003-04

2004-05

2005-06

2006-07

2007-08

2008-09

2009-10

MEAN


TABLE 2: LEVELS OF COMPETITION IN PROCUREMENT

Commodity Type

Quantity Procured

(MTs) Procurement Strategy

Local or Regional Origin

Number of Bids Received

Maize grain 767 Competitive Tender Regional 19 Maize grain 3,464 Competitive Tender Regional 9 Maize grain 468 Competitive Tender Regional 13

Cowpea 76.70 Competitive Tender Local 19 Cowpea 348 Competitive Tender Local 9 Cowpea 46.80 Competitive Tender Local 13

Vegetable Oil 23.89 Limited Competitive Tender Regional 10 Vegetable Oil 123.61 Limited Competitive Tender Regional 10

Salt 1.59 Limited Competitive Tender Regional 10 Salt 8.25 Limited Competitive Tender Regional 10

Millet 430 Competitive Tender Local 19 Millet 50 Limited Competitive Tender Local 3 Millet 100 Limited Competitive Tender Local 1

Source: Mercy Corps Niger Final Evaluation Report excel sheet data; reproduced as it appears in the report, though a note underneath the table confirms that the “Limited Competitive Tenders” of salt and vegetable oil were actually voucher procurements. Also, the “Limited Competitive Tenders” of millet were hard tenders but “invitational,” according to the same note attached to the excel table.


PAKISTAN

WFP LOCAL PROCUREMENT PROJECT, EMERGENCY

In summer 2010, Pakistan experienced monsoon-related flooding that affected more than 20 million people from 78 districts in six provinces. With the USDA LRP funding, WFP Pakistan purchased 10,062 MT of wheat locally from the Government of Pakistan’s Storage and Service Corporation (PASSCO) at a subsidized rate of $276 per MT, which was internationally competitive at the time. The food was distributed in the food deficit areas of Khyber Pakhtunkhwa and Balochistan provinces. WFP food assistance proved critical to maintaining adequate food consumption and protecting the lives of people in distress. In addition to providing food aid, WFP’s project helped beneficiary families to prevent a higher debt burden as they took loans to rebuild after the devastation of the floods. Under the activities supported by this contribution from USDA, improved food consumption among target families remained a key objective.



With a bumper wheat crop prior to the 2010 floods, sizeable wheat stock balances remained available in Pakistan, despite the recurrence of floods in Sindh and Balochistan. As such, and in the context of prevailing economic hardship, the local purchase of wheat by WFP did not impact prevailing market prices––due to the comparatively small scale of transactions––and also helped to ease the domestic market and provide some support to the domestic economy. WFP conducted ongoing in-country market assessments in collaboration with relevant Ministry partners, UN agencies, FEWSNET, NGOs and trader networks, as well as WFP’s food aid monitors, food security and analysis unit, and emergency assessment officers. Prevailing wholesale and retail wheat prices are shown in Table 1, below.


This was a direct purchase from the Government of Pakistan’s stocks, rather than a competitive bid. There was no competition entered into for the procurement of wheat. After consultations with the government through the Ministry of Food and Agriculture, the Ministry confirmed that there were sufficient stocks available to meet WFP requirements from the government’s strategic stocks because there had been a bumper wheat harvest, which the country was trying to export. WFP entered into a long term agreement with PASSCO after lengthy negotiations that aimed to ensure that the price would be lower than that offered by international suppliers.


The area where the wheat was purchased is a wheat surplus area with storage facilities. The market prices of wheat flour were either lower or similar to the prices in the area where food was distributed. The procurement of wheat had no impact on the market in the area of purchase and food was distributed in the food deficit areas of Khyber Pakhtunkhwa and Balochistan provinces. This was because of the comparatively small scale of the procurement and the prior year’s bumper wheat crop. Markets in Multan and Peshawar, two of the four districts where the wheat was distributed, show stable prices over the period prior to and following the procurement (Figure 1). The evaluation team’s findings show no likely impact on consumer prices as a result of the purchase.

III Each government market interference or other activity of the donor country that might have significantly affected the supply or demand of the eligible commodity in the

PASSCO was established in 1973 and became operational in 1974 to support the federal government in ensuring national food security by maintaining strategic reserves of different food grain commodities and providing them to food deficit provinces. At the time of the signing of the agreement with WFP, PASSCO reported a figure of approximately $290 per MT as the cost of the wheat paid to the farmer, excluding the organization’s other costs and overheads. With those other costs, this figure came to $350. The Pakistani government plays a significant role in determining the price paid


Reporting Requirement Notes area at which the local or regional procurement occurred

to wheat farmers. As the largest single purchaser of grain from farmers, the government is positioned to ensure relative price stability within the country. Due to flooding in 2010, this year saw major U.S. in-kind and cash for food aid to Pakistan, including pre-positioning stocks from Djibouti. After the July floods in Pakistan, USAID dispatched pre-positioned peas and vegetable oil from Djibouti and contributed approximately $130 million of Title II and Emergency Food Security Program resources to WFP to ensure food distributions at the start of the floods. With an additional USAID contribution of $90 million at the end of 2010, WFP worked to assist more than seven million flood-affected people. In addition, in 2011 USDA shipped 102,700 MT under the Title II program, 1,200 MT of vegetable oil under the McGovern-Dole International Food for Education and Child Nutrition program and 20,900 MT of vegetable oil under the CCC-funded Food for Progress program.61


WFP Pakistan purchased 10,062 MT of wheat locally from the Government of Pakistan (PASSCO) at a subsidized rate of $276 per MT, internationally competitive at the time, in a direct purchase arrangement.


Procurement took 53 days total, of which seven days were for contracting and 46 were for delivery.


Locally-procured wheat cost $276 per MT for the commodity plus $78.92 per MT for TSH.

ii. Assess: I Whether the requirements

of this section have been met

WFP reporting was found to be compliant with requirements.



The area where the wheat was purchased is a wheat surplus area with storage facilities. The market prices of wheat flour were either lower or similar to the prices in the area where food was distributed. WFP reported that the procurement of wheat had no impact on the market in the area of purchase. The food was distributed in the food deficit areas of Khyber Pakhtunkhwa KPK and Balochistan provinces.

bb Markets The agreement with PASSCO was for the procurement of 300,000 MT of wheat at $276 per MT over a number of months (10,062 MT of which was paid for by USDA LRP project funds). At the time the agreement was signed by WFP and PASSCO, it would have cost WFP $389 per MT to import wheat into Pakistan and clear it; therefore, the price of $276 per MT brought WFP considerable savings. WFP was buying from strategic government stocks, and reported that the purchase had no effect on the demand, supply, or price of wheat. Without detailed data on which to base an independent analysis, the evaluation team could not confirm or refute WFP’s assessment of LRP project impact on wheat prices. Please see Annex 3 and the market impact chapter of the main body of the report for further details.

61 – U.S. Food Aid Tables FY2011 - http://www.fas.usda.gov/excredits/FoodAid/Reports/reports.html


Reporting Requirement Notes cc Low-income consumers The comparatively small scale of procurement minimized market impact (see above);

furthermore, the price was subsidized by PASSCO. Low-income consumers not covered by the program were minimally affected due to the lack of impact on food prices caused by the operation.

dd Program recipients The 10,062 MT of wheat purchased with USDA funds was distributed to 880,425 food-insecure, vulnerable individuals to assist them in meeting their immediate food needs. WFP provided wheat as a portion of the basic food basket through direct distribution. Based on assessment findings, the general family food basket met the majority of an individual’s recommended dietary allowance. Food rations were set to meet the needs of a family of seven. The basic food basket defined by WFP consists of 80kg wheat flour, 8kg pulses, 4.5kg vegetable oil, 1kg salt, 4.5kg High Energy Biscuits (HEB) for children 2-12 years of age, and 1.5kg ready-to-use-supplementary foods (RUSF) for children 6-23 months of age. Beneficiary households were defined on the basis of established vulnerability criteria and from assessment exercise results.

III Time of delivery Procurement took 53 days total, of which seven days were spent on contracting and 46 for delivery.

iii. Compare: Quality and safety

assurances Atlantic Surveyors conducted quality and quantity inspections for WFP on the procurement. Samples were sent to an independent laboratory (Qarshi Research International, Ltd.) and the sample was certified as fit for human consumption.

TABLE 1: WHEAT PRICES, PAKISTAN, JAN-MARCH 2010 Type of Price Month Cost per MT (Pakistani Rupees) Retail January, 2010 PKR 26,580 Retail February, 2010 PKR 26,990 Retail March, 2010 PKR 26,890 Wholesale January, 2010 PKR 23,672 Wholesale February, 2010 PKR 23,875 Wholesale March, 2010 PKR 23,747

Source: WFP, Emergency Food Assistance to Families Affected by Monsoon Floods in Pakistan: Final LRP Pilot Project Report.


FIGURE 1: WHEAT PRICES, PESHAWAR AND MULTAN MARKETS, JUNE 2009-JUNE 2010, IN USD/KG



TANZANIA

WFP LOCAL PROCUREMENT PROJECT, DEVELOPMENT

Through its P4P initiative, WFP used its purchasing power and presence in staple food markets to catalyze innovations that sought to improve market exchange, reduce transaction and marketing costs, and create linkages between smallholder farmers and profitable markets. WFP targeted 29 farmer groups, organized in Savings and Credit Cooperatives (SACCOs), in its LRP project in Tanzania. Under the P4P pilot, a total of $2 million was allocated to purchase from smallholder farmer groups in Tanzania. Between January 2010 and February 2011, 17 contracts were signed for the delivery of 2,475 MT of maize and 424 MT of beans at a total cost of $1,035,708. Maize was procured from ten SACCOs in the northern and central regions of Tanzania (64%), the National Food Reserve Agency (32%), and the Cereal Growers Association of Kongwa (4%). Beans were purchased in the western Kagera region from one SACCO group and one network of SACCO groups (Kaderes).



Smallholder farmers (small-scale subsistence farming) dominate agricultural output and produce virtually all the maize grown in the country. Maize is the most widely grown crop averaging 3-4 million MT per year and is produced by approximately 4.5 million farm households. Maize production is geographically dispersed and maize deficits that negatively affect national food security occur due to supply constraints. Attempting to protect local production, the Government of Tanzania only allows the import of maize as a corrective food security measure when actual demand exceeds domestic supply. Maize prices peaked in 2009 and again during the first half of 2011, reaching record heights in May. The implementation of a six-month export ban caused the subsequent decline in prices. According to the Food FAO, maize prices remained eight percent higher than in 2010 due to high fuel costs and low cereal output (Figure 1).


Maize and beans were procured through competitive tendering and direct purchase. In Tanzania, P4P purchases were mainly conducted through competitive tendering. Direct purchases were also used to encourage new or weaker groups to supply to WFP, to buy when there was an urgent need, or when there were not enough suppliers with stocks to call for a tender. Two direct negotiations in January 2010 resulted in procurements with SACCOs for 277 MT of maize. The deliveries were finalized between March and May 2010. The balance of 2,198 MT was purchased through four competitive tenders, launched in May, August, October, and December 2010. Procurement of beans was conducted through two competitive tenders issued in May and October 2010 for 225 MT. The balance of 199 MT was purchased through direct negotiation.


WFP reported that the percentage difference between minimum and maximum regional reference market prices (such as in Dodoma and Arusha) was less than in markets where P4P purchased. For example, the seasonal highs and lows were greater in the locations where P4P was operating when compared with regional trade hubs. Regional markets, as trading hubs, appeared better integrated with more consistent trade flows, resulting in more stable prices than was found in rural areas where the P4P project was purchasing. Historical price data in rural areas was not available; therefore the evaluation team could not determine whether the seasonal price variation near SACCOs was representative of historically expected behavior. However, considering the relatively small size of the purchases, patterns in seasonal price behavior, and the various WFP procurement activities, purchases from SACCOs do not appear to have exerted upward pressure on wholesale maize prices. Stable wholesale markets likely


Reporting Requirement Notes contributed to similarly stable retail markets. Any deviations observed in retail markets likely did not result from WFP’s particular point of entry into the value chain. From data provided by WFP, the evaluation team concluded that there were no apparent impacts on the price of maize in the P4P markets that were the result of the LRP procurements. Data on bean prices were limited, and as such the evaluation team cannot preclude the possibility of some market impacts for beans. The evaluation team concludes that the possibility exists that there may have been an impact on the price of beans in local markets in the Kagera region due to the LRP purchases of beans. Please see Annex 3 and the market analysis chapter of the report for further details.


WFP reported that although the Tanzanian government is committed to supporting agricultural trade in the region, export bans were imposed following a poor harvest or when consumer prices were unusually high (Table 2). Since 2000, there has been a near-permanent export ban, with only a few months of formal lifting. Maize farmers, traders, and millers have noted, however, that the ban has created unintended secondary effects in terms of lost market opportunities, suppression of domestic prices relative to regional markets, shifting to other crops that are less efficient alternatives under competitive conditions, and diminished investments in trucks and storage capacity.62 WFP also reported that the government imposes no restrictions for entry into the maize or beans production and marketing systems. No licenses are required to start farming or trading in the crops; however, traders who move crops outside of the district are required to pay a statutory fee per bag of commodity (local tax or cess). Local governments set the tax rate at different levels, although the cap is set by the central government at five percent of the maize value.63 In 2010, food aid for Tanzania from the U.S. government included almost 6,000 MT of cornmeal, CSB, green split peas, and vegetable oil under a Title II emergency program; a CCC-funded Food for Progress grant which provided 14,000 MT of wheat; and support from the McGovern Dole International Food for Education and Child Nutrition program of 4,660 MT of sunflower seed, sorghum, rice, and beans.64 In 2011, USAID programmed to ship 7,200 MT of commodities under Title II programming and USDA programmed to ship 14,000 MT of wheat under the Food for Progress program.65


The project purchased 2,425 MT of maize and 424 MT of beans. Maize was procured through two direct negotiations with SACCOs for 277 MT. The balance of 2,198 MT was purchased through four competitive tenders. Procurement of beans was conducted through two competitive tenders issued in May and October 2010 for 225 MT. The balance of 199 MT was purchased through direct negotiation.

62United States Agency for International Development. “AgCLIR: Tanzania. Commercial Legal and Institutional Reform in Tanzania’s Agricultural Sector." USAID Business Climate Legal and Institutional Reform program publication. May 2010. Web, 19 Jun. 2012. http://egateg.usaid.gov/sites/default/files/Tanzania_AgCLIR.pdf 63World Bank. Eastern Africa: A study of the regional maize market and marketing costs. Agriculture and Rural Development Unit (AFTAR). Sustainable Development Department, Country Department 1, Tanzania, Uganda and Burundi, Africa Region, 2009. 64United States Agency for International Development and United States Department of Agriculture. U.S. International Food Assistance Report 2010. N.d. Web. 14 Apr. 2010. http://www.usaid.gov/our_work/humanitarian_assistance/ffp/fy2010.ifarreport.pdf. 65 U.S. Food Aid Tables FY2011, <http://www.fas.usda.gov/excredits/FoodAid/Reports/reports.html>.


Reporting Requirement Notes V Timeframe of each

procurement of each eligible commodity

Average overall times for procurement were as follows: • Beans averaged 153 days to procure, with 37 days for contracting and 116

days for delivery66 and • Maize required 132 days to procure, with 50 days for contracting and 82 days

for delivery. Note that these times reflect the capacity building development goals of the project.

VI Total cost of procurement including storage, handling, transport, and administrative costs

Note that in Tanzania, TSH was included in commodity costs for a few procurements (Delivered Duty Unpaid or DDU). At the request of the evaluation team, WFP Tanzania estimated the portion of commodity costs that would otherwise have covered TSH costs, which was calculated by subtracting the selected supplier's Free Carrier (FCA) price from its Delivered at Place (DAP) price. The result of this calculation (carried out by WFP and submitted to the evaluation team) was then subtracted from the commodity cost and placed in the dataset as TSH. Average MT costs for hard tender procurements were $566.77/MT for beans plus $104/MT for TSH. For maize grain, the average costs were $243.99/MT plus $33.63/MT for TSH. For direct purchases, average costs for beans were $610.93/MT plus $143.44/MT for TSH, while for maize, average costs were $276.14/MT plus $46.46/MT for TSH.

ii. Assess: I Whether the

requirements of this section have been met

WFP reporting was found to be compliant with requirements.



WFP encouraged P4P-assisted FOs to engage in collective marketing to achieve economies of scale and reduce transaction costs. Village warehouses were rehabilitated and fully equipped to facilitate storage of commodities and maintenance, ensuring good quality. P4P also provided connections with financial institutions to provide credit facilities for aggregation. These improvements to the SACCOs’ capacity are significant for SACCOs’ competitiveness and the ability of these organizations to bring commodities to market, whether for development, emergency or commercial purposes.

bb Markets With regard to beans, WFP reported that given the small percentage of the commodity purchased out of the total production in the area (0.1%), the risk was low for any negative impact on local consumers. From data provided by WFP, the evaluation team concluded that there were no apparent impacts on the price of maize in the P4P markets that were the result of the LRP procurements. Data on bean prices were limited, and as such the evaluation team cannot preclude the possibility of some market impacts for beans. The evaluation team concludes that the possibility exists that there may have been an impact on the price of beans in local markets in the Kagera region due to the LRP purchases of beans. Please see Annex 3 and the market analysis chapter of the report for further details.

cc Low-income consumers Commodities were distributed between December 2010 and June 2011 under refugee and HIV/AIDS programs, including to patients on Anti-Retroviral Treatment (ART), women involved in the Prevention of Mother-to-Child Transmission (PMTCT) program, and the Home Based Care (HBC)/Orphans program. USDA food reached a

66 Price fluctuations prevented farmers’ associations from complying with contracts on the schedule intended, according to WFP Tanzania. “Contracts were signed with Farmers Groups for immediate deliveries…. [A]t the time of deliveries, market prices had increased and Farmers Groups were then no longer in a position to deliver to WFP since the contracted prices were lower prices than the current market prices. WFP did not cancel the contracts, which remained in force. It is only at the onset of the new crops season when the market prices matched the WFP contracted prices that the Farmers managed to deliver.”


Reporting Requirement Notes total of 99,591 refugees: 4,939 beneficiaries under the HBC/Orphans program and 728 ART/PMTCT patients.67

dd Program recipients WFP encouraged P4P-assisted FOs to engage in collective marketing to achieve economies of scale and reduce transaction costs. Village warehouses were rehabilitated and fully equipped to facilitate storage of commodities and maintenance of good quality. P4P also provided connections with financial institutions to provide credit facilities for aggregation. Beneficiaries of the food aid, primarily groups with significant nutritional needs, improved their food security through the project as well.

III Time of delivery Average overall times for procurement were as follows: • Beans averaged 153 days to procure, with 37 days for contracting and 116

days for delivery and • Maize required 132 days to procure, with 50 days for contracting and 82 days

for delivery. Note that these times reflect the capacity building development goals of the project.

iii. Compare:


WFP reported that the quality of maize and beans received from all SACCOs was satisfactory and met required standards. WFP reported that they supervised the process of cleaning and re-bagging to ensure required quality standards were met and the same standards were applied when buying from regular traders. African Marine Surveyors and Consultants, Ltd. tested the WFP Tanzania commodities for aflatoxin. In addition, they conducted a range of other standard tests and found no safety or quality problems with the commodity samples.

67 USDA Local and Regional Procurement Pilot Project, Evaluation Report –Final LRP Project Report World Food Programme/Tanzania. Tanzania: WFP Tanzania Country Office and Research on Poverty Alleviation, 2011.


FIGURE 2: MAIZE PRICES (WHOLESALE) NOVEMBER 2007 TO NOVEMBER 2011

Source: FAO GIEWS Food Price Data and Analysis Tool at http://www.fao.org/giews/pricetool2


UGANDA

WORLD VISION LOCAL PROCUREMENT PROJECT, DEVELOPMENT

World Vision’s Acholi Voucher Project (AVP) was implemented in the districts of Pader and Kitgum in the Acholi sub-region of northern Uganda. A total of 11 sub-counties, five in Kitgum and six in Pader/Agago, were identified for assistance based on their categorization as the most food-insecure areas by a 2009 Food and Agriculture Organization Rapid Food Security Assessment.

The Acholi region is in the process of recovering from nearly two decades of war. In the past five years, markets have been reopening and production has increased; however, it has only been in the past two years that a concerted effort has been made to move the internally displaced persons (IDPs) out of camps and back to their homelands. Markets in the region have grown quickly as the IDPs have begun to return. Because of the functioning markets, WV was able to use a voucher-based distribution approach and directed some food for work project activities at opening roads to markets.

The project provided food through vouchers to 51,111 beneficiaries. Vouchers were provided for maize, beans, and cooking oil. One hundred and sixty-two vendors took part in the project and approximately 2,071 MT of cereals (both milled and unprocessed), 874 MT of beans, and 180 MT of cooking oil were provided via voucher distribution to beneficiaries.


i. For each market & commodity procured, describe: I Prevailing and historic


Uganda produced just below 1.3 million MT of maize in 2010. 68 Annual maize production has been relatively stable between 1 and 1.4 million MT since 2000, showing a slightly rising trend each year (Figure 1). Uganda produced almost 460,000 MT of beans in 2010. Dry bean production fluctuated from 420,000 MT to 535,000 MT between 2000 and 2010, trending downward over this period (Figure 2). Vegetable oil production approximately doubled from 72,000 MT in 2000 to 142,000 MT in 2010. Sunflower, soybean, and sesame are the oils most commonly produced (Figure 3). Data on historical demand in the country is sparse. A food balance sheet from 2007 is provided in Table 1 below, but more recent secondary data were not available. Real wholesale maize prices between 2006 and 2011 indicated a rising trend with noticeable price spikes in 2008 due to the global food price crisis, and in early 2011. Available wholesale maize price data for Lira showed similar price movements to those seen in Kampala markets, suggesting these markets are closely integrated (Figure 4). Real wholesale bean prices from 2006 to 2011 also indicated a rising price trend. Available Lira data also showed similar price levels and movements compared to those seen in Kampala (Figure 5).

68 Analysis from data obtained through the UN Food and Agriculture Organization’s FAOSTAT website. http://faostat.fao.org/


Reporting Requirement Notes Maize prices increased in the April to August period in 2008, 2009, and 2011. This is consistent with the ‘hunger period’ that usually occurs from April to August in Uganda. Seasonal patterns for 2007 and 2011 showed prices gradually falling as the year progressed (Figure 6). Bean prices gradually increased from January until October or November (Figure 7). Dips in price were seen in June to July in most years. 2011 was unusual as wholesale bean prices reached a peak in May and June after which they consistently declined. This is somewhat similar to 2011 maize prices which peaked in July and then fell.


World Vision informed potential suppliers through community meetings, newspaper advertisements and market visits. Vendors were approved to participate in the voucher program on the basis of storage capacity, quality and quantity of goods, possession of local business registration permits, and possession of a scale for use in voucher exchange. 162 vendors were selected to participate and received an orientation and basic training in commodity handling and bookkeeping. Approved vendors were monitored regularly for compliance, and 24 were terminated over the duration of the project for not meeting quality or redemption standards.


World Vision reported that the distribution of commodities via a voucher program had no economically or statistically significant effect on maize grain or sorghum prices for consumers in Uganda in the month of, or following, voucher distribution. However, the dates registered for market price data do not coincide with the dates those goods were purchased from producers to be sold in the vendor scheme. The date at which those purchases were made is the date on which producer or source market prices might have been affected. As a result, the evaluation team determined that it was not possible to say whether World Vision’s voucher project caused any real price impact at the producer level. World Vision did report that voucher distributions were associated with statistically significant, temporary decreases in local market prices of maize flour and dried beans.


World Vision reported no government market interference during the course of the project. Country visits by evaluation team members also found no government interference in market supply or demand. The U.S. Government provided over 15,500 MT of cereals, pulses, oils, and FBFs to Uganda in 2010 under a Title II emergency program, and over 30,000 MT of commodities through a Title II development program, plus 8,400 MT through the McGovern Dole International Food for Education and Child Nutrition program.69 USAID programmed 33,400 MT through a Title II program, and USDA programmed 4,500 MT of commodities to be shipped through the McGovern- Dole International Food for Education and Child Nutrition program for 2011.70


Approximately 2,070.79 MT of cereals, 874.88 MT of pulses, and 179.73 MT of cooking oil were distributed to beneficiaries through the use of voucher exchange. Please note that cereals procured through this project fall under both milled and unprocessed categories, with 1,356.78 MT of unprocessed cereals and 714.01 MT of milled cereals procured overall.


The voucher distribution dates are shown below in Table 2. Beneficiaries were instructed to redeem the vouchers within 20 days of distribution. World Vision reported that most beneficiaries redeemed their vouchers within 10 days. Overall, 15 days elapsed for the entire voucher procurement process, including contracting and delivery times.

69United States Agency for International Development and United States Department of Agriculture. U.S. International Food Assistance Report 2010. N.d. Web. 14 Apr. 2010. <http://www.usaid.gov/our_work/humanitarian_assistance/ffp/fy2010.ifarreport.pdf>. 70U,S, Food Aid Tables FY2011, <http://www.fas.usda.gov/excredits/FoodAid/Reports/reports.html>.




Average costs per commodity were as follows: • Unprocessed cereals cost $246.66 per MT; • Milled cereals cost $517.74 per MT; • Beans cost $741.11 per MT; and • Vegetable oil cost $1,575.00 per MT.

All TSH costs were $1.91 per MT for all commodities procured.

ii. Assess:


World Vision reporting was found to be compliant with requirements.



World Vision reported that due to a demand for a wide variety of food commodities from vouchers, local agricultural producers now have a ready market for a wide range of commodities. World Vision further reported that local farmers are likely to get better prices from buyers due to the improved quality of their goods, which was required by the LRP project, as well as improved knowledge of regional prices due to exposure to many sellers in the marketplace. With this improved price transparency, local farmers now have a better bargaining position when establishing prices with vendors and traders. Local farmers were usually small-scale (subsistence or semi-commercial) farmers. No specific information was found as part of this study related to large-scale (commercial) agricultural producers.

bb Markets

The LRP project allowed registered vendors to increase profitability, strengthening their business. Some vendors opened outlets near communities that received vouchers and closed the outlets when the program finished. The LRP project has increased vendor capacity. This was reported by World Vision and related by vendors during evaluation fieldwork. Certified vendors were able to open bank accounts, thus increasing savings and gaining access to credit. Vendors’ capacity regarding commodity quality, safety, storage, record keeping, reconciliation, banking, tendering procedures, and the establishment of clean storage facilities was developed. Links were created between local vendors and regional commodity wholesalers, with local vendors selling commodities at lower prices and regional wholesalers benefitting from an increase in trade volume.

cc Low-income consumers The short-term decreases in prices for maize flour and dried beans as reported by World Vision probably benefited low-income consumers. Non-beneficiary low-income consumers interviewed during the evaluation team’s fieldwork generally wanted to join future voucher programs, though they did not report specific price volatility. World Vision reported that low-income consumers were able to see which vendors were identified as voucher registrants. With this identification came the understanding that the vendors were held to higher standards of quality and professionalism.

dd Program recipients The evaluation team found program recipients were able to use vouchers to access a wide variety of foods. They were also able to save money for other uses such as paying their children’s school fees, or reducing the amount of piecework they performed. Beneficiaries noted that the project increased attention to the quality of commodities, cleanliness of business premises, and other hygiene and quality


Reporting Requirement Notes testing procedures on the part of vendors. Beneficiaries’ exposure to a number of vendors also improved their knowledge of local prices, allowing them to plan their purchases accordingly. 51,111 individuals from 9,265 families benefited from commodities provided through the AVP.

III Time of delivery Beneficiaries were instructed to redeem the vouchers within 20 days of distribution. World Vision reported that most beneficiaries redeemed their vouchers within 10 days; overall, 15 days elapsed for the entire voucher procurement process.

iii. Compare:

IV Quality and safety assurances World Vision reported that laboratory analysis results showed most commodities passed the majority of the tests with one or two exceptions. Out of all the samples collected over the project’s history, only one sample tested positive for aflatoxin, and even in that case the levels were within the acceptable range. Based on recommendations from the laboratories conducting the analyses, the food was rated good for human consumption even though deviations from the standard parameters were observed in moisture content and foreign matter.

FIGURE 1: MAIZE PRODUCTION, UGANDA, 2000-2010, IN MT


0

200000

400000

600000

800000

1000000

1200000

1400000

1600000

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Met

ric T

ons


FIGURE 2: BEAN PRODUCTION, UGANDA, 2000-2010, IN MT


FIGURE 3: VEGETABLE OIL PRODUCTION, UGANDA, 2000-2010, IN MT


0

100000

200000

300000

400000

500000

600000

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Met

ric T

ons

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 -

20,000

40,000

60,000

80,000

100,000

120,000

140,000

160,000

Sunflower oil

Soybean oil

Sesame oil

Groundnut oil

Cottonseed oil

MT


FIGURE 4: WHOLESALE MAIZE PRICES, UGANDA, 2001 TO 2007, IN USD/MT


FIGURE 5: WHOLESALE BEAN PRICES, UGANDA, 2001 TO 2007, IN USD/MT


0

50

100

150

200

250

300

350

400

450

Dec-

01

Apr-

02

Aug-

02

Dec-

02

Apr-

03

Aug-

03

Dec-

03

Apr-

04

Aug-

04

Dec-

04

Apr-

05

Aug-

05

Dec-

05

Apr-

06

Aug-

06

Dec-

06

Apr-

07

Aug-

07

Who

lesa

le (U

S$ re

al /

tonn

e)

Kampala

Mbale

Lira

Linear (Kampala )

0

100

200

300

400

500

600

700

800

Dec-

01

Apr-

02

Aug-

02

Dec-

02

Apr-

03

Aug-

03

Dec-

03

Apr-

04

Aug-

04

Dec-

04

Apr-

05

Aug-

05

Dec-

05

Apr-

06

Aug-

06

Dec-

06

Apr-

07

Aug-

07

Who

lesa

le (U

S$ re

al /

tonn

e)

Kampala

Lira

Linear (Kampala)


FIGURE 6: WHOLESALE MAIZE PRICES, KAMPALA, UGANDA, 2007 TO 2011, IN USD/MT


FIGURE 7: WHOLESALE BEAN PRICES, KAMPALA, UGANDA, 2007 TO 2011, IN

USD/MT


0

50

100

150

200

250

300

350

400

450

Jan Feb Mar Ap May June July Aug Sep Oct Nov Dec

Who

lesa

le (U

S$ re

al /

tonn

e)

2007

2008

2009

2010

2011

Average

0

100

200

300

400

500

600

700

800

Jan Feb Mar Ap May June July Aug Sep Oct Nov Dec

Who

lesa

le (U

S$ re

al /

tonn

e)

2007

2008

2009

2010

2011

Average


TABLE 1: UGANDA FOOD BALANCE SHEET 2007 (‘000 MT)

Food Flows Maize Cassava Millet Sorghum Beans

Total Production 1,262 4,973 732 456 435

Imports (+) 43 0 0 78 0

Exports (-) 107 3 0 0 8

Total Consumption 893 2,920 590 230 405


TABLE 2: VOUCHER REDEMPTION RATES

Voucher Distributed Vouchers Redeemed Difference Percent Achievement

Pader/Agago 2,400,515,500.000 2,229,593,898.600 170,921,601.40 93%

Kitgum 1,875,312,500.000 1,843,535,943.640 31,776,556.36 98%

Total 4,275,828,000.00 4,073,129,842.24 202,698,157.76 95% Source: USDA Local and Regional Procurement Pilot Project, Evaluation Report –Final LRP Project Report. World Vision/Uganda


ZAMBIA

LAND O’LAKES LOCAL PROCUREMENT PROJECT, DEVELOPMENT

USDA awarded Land O'Lakes $3,624,017 to implement the Zambia Local and Regional Procurement project (ZLRP). The goal of ZLRP was to provide nutritious, safe, timely, and locally procured food to 10,000 food-insecure households. These households were targeted by the project to support the nutritional needs of people living with HIV and AIDS, and children under the age of five, in parts of Zambia with a higher than average prevalence of HIV and child stunting. The project purchased 2,745 MT of fortified maize meal, 124 MT of maize grain, 500 MT of beans, 173 MT of vegetable oil, and 50 MT of High Energy Protein Supplement (HEPS) through competitive tenders via the Zambia Agricultural Commodity Exchange (ZAMACE) in Lusaka, Zambia. The program also purchased 87 MT of HEPS via direct purchases from the local organization, Community Markets for Conservation (COMACO), founded by the Wildlife Conservation Society (WCS). Finally, 1,126 households received hammer mill vouchers, enabling beneficiaries to mill maize grain by 18 hammer mills in four final distribution points. These households received 110 kg bags of maize grain, which provided 100 kg of maize meal once milled. All commodities were produced in Zambia and distributed within Zambia, in three districts: Chongwe, Chibombo, and Mumbwa.


i. For each market & commodity procured, describe:


Maize production figures show significant growth in production levels since 2009 (Table 1). This was due to favorable climatic conditions and government initiatives, such as one (the Fertilizer Input Supply Program, or FISP) that made fertilizers accessible to farmers. Another government initiative that provided incentives was the food purchases made by the government’s Food Reserve Agency (FRA). Pulse production averages only about 24,000 MT annually. Trends in production increased (Table 2) with good harvests in 2009 and 2010. Dried beans are mainly bought by local households and on a relatively large scale by the WFP from the ZAMACE. Zambia has seen a considerable surge in soybean and sesame oil production in recent years. The present consumption of vegetable oil is about 72,000 MT, while local production is about 30,000 MT, with imports making up the balance. Maize prices in Zambia show a rising historical trend (Figure 1) between 1999 and 2007. Maize prices peaked in 2005. A second peak also occurred from 2008 to 2009 in line with the global food price crisis. Since 2009, Zambian maize prices have fallen due to favorable weather and government intervention. Dried bean prices are shown in Figure 2. Price movements show little variation over the time period presented (2006 to 2010). An exception is 2008 when prices were above average, which is consistent with the timing of globally higher food prices. Vegetable oil prices for 2006 to 2010 are shown in Figure 3. Again, historical price movements show highest prices in 2008.


Levels of competition in procurement bids are presented in Table 3. Three procurements of HEPS were single-sourced, and the remaining thirteen


Reporting Requirement Notes procurements for all commodities were tendered through ZAMACE. Of the thirteen competitively sourced procurements, three received only one bid; the remainder received two or more bids each. ZAMACE is a fairly new exchange system and Land O’Lakes’ project tendered two commodities on the exchange that had not been sold there before – vegetable oil and HEPS. Land O’Lakes also attempted to increase bids by publishing the tender announcements widely prior to the release of the tenders, which allowed interested suppliers to partner with a ZAMACE member and place bids. Land O’Lakes purchased 88.35 MT of HEPS from WCS/COMACO on a no-bid, single-source direct purchase contract. Land O’ Lakes used this procurement approach to support the development of a medium-scale community-based processor.


Given the small scale of procurement through the LRP Project relative to national production, impact on national producer and consumer prices is seen as likely to be minimal (Table 4). Land O’Lakes' local purchases of beans under the USDA LRP Project had no economically or statistically significant effect on retail bean prices in Zambia. During the period of the ZLRP project, the FRA was actively engaged in the maize market. The FRA purchased a total of 883,036 MT of maize during 2010 and 2011. Reportedly, it purchased maize grain in the market and then sold it at a sharp (approximately 30%) discount to millers. As part of the Learning Alliance, Cornell University conducted an analysis on whether the ZLRP had an impact on market prices. As part of this analysis, Cornell University researched the FRA’s purchases, but could not find adequate data on the FRA’s purchases and sales and thus was unable to control for this intervention activity. Cornell University detected small price decreases, but estimated that they could not be attributable to LRP. The evaluation team also tried to source FRA purchase and sales data, but was unsuccessful. It is highly possible that Cornell University’s findings reflect not only LRP activity but also the FRA’s interventions. The volumes of FRA interventions are significant and are very likely to affect maize market prices. This supports Cornell University’s statement that they believe that it is unlikely that Land O’Lakes’ activities alone were fully responsible for the maize meal procurement and distribution price effects, and it is likely that other, contemporaneous factors for which researchers cannot adequately control also influenced these prices. As a result of these analyses, the evaluation team finds price impact from this LRP project unlikely. Please see Annex 3 and the market impact chapter of the main body of the report for further detail.


The Zambian Government engages in a high level of interference in supply and demand. FISP provides fertilizer and improved maize seed to farmers in an effort to stimulate the supply of maize. One the demand side, a major factor in Zambia’s maize markets is the FRA. Initially tasked with holding strategic grain reserves, the FRA has become a major maize buyer. In 2010 and 2011, the FRA offered to buy unlimited quantities of maize at an attractive price of ZMK 65,000/50kg (approx. $260/MT). FRA wholesale purchase prices have been above retail selling prices


Reporting Requirement Notes since 2009/2010. Selling to the FRA is made more attractive to smallholder farmers, as purchases are usually made locally. USAID distributed bulgur wheat through its MYAP called the Consortium for Agriculture and Nutrition, AIDS, Resiliency and Marketing (CFAARM) project. This project operated in southern and western Zambia, not near LRP target areas. WFP also had a P4P initiative that tried to link producers with private sector buyers in order to increase maize purchases. These initiatives were reported to be of very small scale.


500.48 MT of beans, 2,745 MT of maize meal, 124.4 MT of maize grain, 173.86 MT of vegetable oil, and 50.15 MT of HEPS were procured via hard tender. 87.91 MT of HEPS were procured via direct purchase.


Total average times for the 16 procurements were as follows: Maize meal took a total of 49 days to procure, with 15 days for contracting, and 33 days for delivery. Maize grain took a total of 68 days to procure, with 12 days for contracting and 56 days for delivery. Beans took a total of 49 days to procure, with 26 days for contracting and 33 days for delivery. Vegetable oil took a total of 49 days to procure, with 26 days for contracting and 24 days for delivery. HEPS procured via hard tender took a total of 44 days, with 16 days for contracting and 28 days for delivery. HEPS via direct purchases took a total of 91 days to procure, with 69 days for contracting and 22 days for delivery. At the end of the project, a supplier default affected the pace of distribution and the ration sizes received by a number of households. As a result, about 400 households still received a monthly distribution, but received a smaller ration of maize meal. As the default occurred at the end of the project, Land O’Lakes did not have time to source the commodities from another supplier.


Average costs per MT were as follows: Hard tenders: Beans cost $1,075.61 per MT, plus $235.25 per MT for TSH costs; HEPS cost $895.10 per MT, plus $194.40 per MT for TSH; Maize meal cost $198.87 per MT, plus $135.92 per MT for TSH; Maize grain cost $226.00 per MT for commodities, plus $242.37 per MT for TSH; and Vegetable oil cost $1,682.26 per MT for commodities, plus $368.11 per MT for TSH. Direct purchase: HEPS cost $729.38 per MT plus $246.77 per MT for TSH.

ii. Assess:


Land O’Lakes’ reporting was found to be compliant with requirements.





The evaluation team found no evidence of impacts on local large-scale producers. However, during the evaluation site visits, one Lusaka-based bean supplier noted that the smallholder farmer groups which supplied the company were pleased with the additional demand for beans. Hammer millers in distribution areas were concerned that the provision of maize meal to beneficiaries had a negative effect on their business; Land O’Lakes responded with vouchers for use at the hammer mills and the provision of maize grain to beneficiaries, to be milled by the approved millers.

bb Markets

Key informant interviews by the evaluation team suggested some large-scale processors, traders, brokers, and the commodity exchange did benefit from the opportunities that supplying the LRP project presented. The volumes purchased through the LRP project stimulated business for those who won or facilitated contracts. In addition, the ZAMACE exchange was in the process of restructuring in order to garner participation from a more diverse array of suppliers. The LRP project supported this goal and widened the exposure of the exchange to an additional sector of the donor market. Local-level hammer millers reported negative impacts from the LRP project. The evaluation team interviews with hammer millers showed that distribution of maize meal reduced business for hammer millers as their customers no longer needed to mill maize. When Land O’Lakes became aware of this, they distributed maize grain with milling vouchers, so that beneficiaries would continue to use the milling services, in an attempt to overcome the issue. However, hammer millers reported that vouchers were not reimbursed in a timely manner, delaying cash flow. Millers could not buy fuel or pay for machine maintenance. Some millers reported that this stopped maize grinding or reduced turnover. Evaluation team interviews with traders found some small-scale shopkeepers were positively affected by the LRP project. Recipients of the food aid rations under the LRP project were reported to have saved money, allowing more spending on ‘luxury’ products such as vegetables and soap. Some market vendors also reported a reduction in the number of traders visiting local markets to sell commodities that were being supplied by the LRP project. These traders returned once the program ended.

cc Low-income consumers Given the small scale of LRP purchases, and little or no price volatility for beans, the evaluation team finds price impact, and thereby impact on low-income consumers, from this LRP project unlikely. Please see Annex 3 and the market impact chapter of the main body of the report for further detail.

dd Program recipients Evaluation interviews with beneficiaries showed that LRP recipients saved money that they would otherwise have used to buy food, or spent on other items. This money was regularly reported as having been used for school fees; vendors in markets nearby said there was an increase in purchases of items like soap and vegetables. Recipients also reported reducing the amount of piecework they performed to earn money. Land O’Lakes’ subcontractor on the agreement, World Vision, carried out an end-of-project evaluation in September 2011, in which 58 percent of beneficiaries interviewed stated that they were consuming more and different types of food, 20 percent said their children were eating more food than before, 27.4 percent said they spent more on food, and 25.4 percent stated they spent less time on working other peoples’ farms.


Reporting Requirement Notes Between May and August 2011, the incidence of severe malnutrition declined from 9.3 percent to 5.8 percent, while the incidence of moderate malnutrition fell from 17.39 percent to 12.29 percent among beneficiaries. Over 60,100 members of households affected by HIV/AIDS received rations. Some 8,600 young children (from 6-24 months old) received supplemental rations.

III Time of delivery Total average times for the 16 procurements were as follows:

• Maize meal took a total of 49 days to procure, with 15 days for contracting, and 33 days for delivery;

• Maize grain took a total of 68 days to procure, with 12 days for contracting and 56 days for delivery;

• Beans took a total of 49 days to procure, with 26 days for contracting and 33 days for delivery;

• Vegetable oil took a total of 49 days to procure, with 26 days for contracting and 24 days for delivery; and

• HEPS procured via hard tender took a total of 44 days, with 16 days for contracting and 28 days for delivery. HEPS via direct purchases took a total of 91 days to procure, with 69 days for contracting and 22 days for delivery.

At the end of the project, a supplier default affected the pace of distribution and the ration sizes received by a number of households. As a result, about 400 households still received a monthly distribution, but received a smaller ration of maize meal. As the default occurred at the end of the project, Land O’Lakes did not have time to source the commodities from another supplier.

iii. Compare:

IV Quality and safety assurances ZLRP took a number of actions to ensure the quality of the commodities it received. First, ZLRP launched a competitive tender for commodity quality and ingredients testing which was awarded to SOCOTEC. SOCOTEC tested all commodities to ensure that they met contract specifications, and that the total quantity contracted was received. Land O’Lakes did not take possession of commodities that did not meet specifications. Sales contracts ensured that processors and vendors who produced sub-standard commodities, or delivered less than the contracted quantity, replaced the rejected amount, or fulfilled the shortfall. Commodity vendors complied with requests to replace commodities, anywhere from one week to one month after the project rejected the commodity. At the same time, ZLRP focused attention on building processor capacity, including working with COMACO to build its efficiency and production quality and conducting food safety trainings with the Zambian Bureau of Statistics. Specific training activities focused on Good Manufacturing Practices (GMP) and basic Hazard Analysis and Critical Control Points (HACCP) for 14 millers and processors, with additional training offered on similar topics to 18 local hammer millers. Despite the measures taken by Land O’Lakes, two out of three hammer millers interviewed by the evaluation team reported that some maize grain was contaminated with metal and stones, which damaged their equipment and led to losses.


Reporting Requirement Notes Land O’Lakes bought a lower quality maize meal similar to roller meal as opposed to the higher quality breakfast meal since it was more nutritious. Some beneficiaries mentioned the ‘bitter’ taste usually associated with roller meal. Quality testing laboratory SOCOTEC reported that the initial fat content was a problem with the maize meal, but that fat content decreased in subsequent deliveries. No problems were reported with micronutrient or aflatoxin levels. SOCOTEC reported that testing should have taken five days, but lab queues led to this being exceeded. HEPS and maize meal were the most delayed as an external lab was needed to test these commodities.

TABLE 1: MAIZE PRODUCTION (MT)

2002 2003 2004 2005 2006 2007 2008 2009 2010

Maize 606,172 1,157,860 1,214,000 866,187 1,424,400 1,366,160 1,211,570 1,887,010 2,795,480


TABLE 2: PULSE PRODUCTION (MT)

2002 2003 2004 2005 2006 2007 2008 2009 2010

Pulses 19,000 20,500 22,675 22,500 23,000 23,500 23,911 26,435 27,000


FIGURE 1: MAIZE PRICES


0

50

100

150

200

250

300

350

400

450

Dec-

99

Apr-

00

Aug-

00

Dec-

00

Apr-

01

Aug-

01

Dec-

01

Apr-

02

Aug-

02

Dec-

02

Apr-

03

Aug-

03

Dec-

03

Apr-

04

Aug-

04

Dec-

04

Apr-

05

Aug-

05

Dec-

05

Apr-

06

Aug-

06

Dec-

06

Apr-

07

Aug-

07

Dec-

07

Reta

il N

atio

nal A

vera

ge (U

S$ /

tonn

e)


FIGURE 2: DRIED BEAN PRICES $/KG 2006-2010

Source: USDA Local and Regional Procurement Pilot Project, Evaluation Report Final LRP Project Report Land O’Lakes/Zambia: Questions/Further Clarification Items

FIGURE 3: VEGETABLE OIL PRICES $/KG 2006-2010

Source: USDA Local and Regional Procurement Pilot Project, Evaluation Report Final LRP Project Report Land O’Lakes/Zambia: Questions/Further Clarification Items


TABLE 3: TENDERS, BIDS, AND CONTRACTS

Commodity Quantity Procured

(MTs) Procurement Strategy

Number of Bids

Received

Date Contract Signed (day/month/year)

Maize meal 1800.0 Competitive tender on exchange 4 22/12/2010

Beans 250.15 Competitive tender on exchange 3 22/12/2010

Vegetable oil 100.02 Competitive tender on exchange 4 22/12/2010

HEPS 12.5 Competitive tender on exchange 2 22/12/2010



HEPS 38.32 Non‐competitive, single sourcing 1 17/05/2011

Maize meal 738.0 Competitive tender on exchange 2 01/06/2011



Maize meal 443.7 Non‐competitive, single sourcing 1 16/08/2011

Maize grain 124.5 Competitive tender on exchange 2 1/8/2011




HEPS 25.0 Non‐competitive, single sourcing 1 26/08/2011

Source: USDA Local and Regional Procurement Pilot Project, Evaluation Report –Final LRP Project Report Land O’Lakes/Zambia

TABLE 4: COMPARISON OF LRP VOLUMES TO NATIONAL PRODUCTION

Commodities Local Production (‘000 MT)

Purchased for Distribution (‘000 MT)71

Percent of Total Production Purchased

Maize grain (2011) 2,800 5.00 0.18%

Vegetable oil (5-year average) 24 0.18 0.75%

Beans (5-year average) 30 0.50 1.67%

Source: USDA Local and Regional Procurement Pilot Project, Evaluation Report –Final LRP Project Report Land O’Lakes/Zambia 71 Maize meal purchase assumes amounts required for maize grain, to produce maize meal.


ZIMBABWE

UNITED METHODIST COMMITTEE ON RELIEF REGIONAL PROCUREMENT PROJECT, EMERGENCY

The United Methodist Committee on Relief (UMCOR) received an award to implement an emergency food aid distribution project in Chipinge District, Zimbabwe. The Zimbabwe pilot project procured eligible food commodities—non-GMO maize, yellow peas, and vitamin A-fortified cooking oil—regionally from South Africa. The objective of UMCOR’s emergency response project was to ensure adequate short-term food availability for seasonally and chronically food-insecure households affected by drought conditions and economic shocks. UMCOR served over 93,000 individually packaged rations to more than 68,000 beneficiaries during three implementation phases from March to August, 2011. Each phase reached a unique set of beneficiaries. UMCOR categorized the beneficiaries into two groups: transitorily food-insecure households, and chronically food-insecure households. UMCOR subcontracted Nathan Associates, Inc. to conduct a pre-procurement price analysis, monitor market prices during the project, and produce the final market monitoring report.

Reporting Requirement Notes i. For each market & commodity procured, describe:


Zimbabwean production levels for maize are volatile and show a declining trend over time (Figure 1). Reasons for the downward trend include a gradual switch by large-scale commercial farms from maize to cash crops, principally tobacco and cotton. Other government interventions and economic events have also contributed to the decline: land tenure policies, lack of investments in agriculture, and economic deterioration. A large proportion of the maize produced in the country comes from small-scale farmers, whose average yields are a fraction of those of larger producers elsewhere in the region. Low productivity is often due to limited knowledge, especially among the newly resettled communal farmers, as well as the unreliable supply of low-cost inputs (primarily seed, fertilizer, and electricity). An input support program increased the availability of inputs and maize production did increase over the last two agricultural seasons. Maize is the staple food in Zimbabwe and is usually consumed in the form of flour (mealie meal). Upper and lower annual consumption estimates, respectively, are 1.88 million to 1.05 million MT, but a reasonable estimate, based upon a population of 10.75 million people and an annual per capita consumption of 140 kg, is 1.51 million MT. An additional 100,000 MT above the current level of production is required for stock feed, resulting in a total national demand estimated at 1.61 million MT. Maize prices in Zimbabwe fluctuated from $0.05/kg to $1.15/kg between 2005 and 2009. From July 2008 to December 2008, price levels increased significantly, corresponding with increases in global cereal prices. Since April 2009, prices have been relatively stable (between $0.23/kg and $0.34/kg). However, Zimbabwe’s switch in 2009 to the U.S. dollar has brought an unexpected volatility to retail prices, as the country lacks sufficient coins for consumers to use. Price increases in rural areas are more commonly in $1 increments, while the government attempts to find a solution to the lack of coins. In some parts of the country, the use of the South African rand in place of U.S. coins developed. Historic prices for maize in South Africa, where UMCOR’s purchases were made, are presented in Figure 2.


Reporting Requirement Notes Zimbabwe’s cooking oil consumption was estimated at about 100,000 MT per year. Oil consumption was mostly in the form of blended vegetable oils with a high proportion of imports coming from South Africa (mostly sunflower oil). Zimbabwean retail prices of imported vegetable oil have been fairly stable since the dollarization of the economy in 2009. Yellow peas are rarely produced or consumed in Zimbabwe.


All commodities were procured via hard tender under this project. Using a list of WFP-approved vendors in South Africa, UMCOR received seven bids for the maize tender, five bids for the yellow peas, and three bids for the vegetable oil.


UMCOR reported that it purchased 1,291 MT of white maize in Gauteng Province, South Africa. A total of 685,000 MT of white maize was produced in Gauteng Province in 2009-2010, making the UMCOR procurement approximately 0.2 percent of total maize production in the province at the time. This strongly suggests that, at the South African regional level, no market price impacts would have been likely in Gauteng Province as a result of this procurement. Since the national (South African) harvest was over 7.8 million MT, the relatively small LRP purchase would not have had any discernible price or supply impact at the national level. The evaluation team found no visually obvious spikes or decreases in the price of white maize in the data around the time of the procurement. UMCOR purchased 89 MT of cooking oil, representing 0.037 percent of the national South African production of 242,805 MT of this commodity. UMCOR purchased 1.2 percent of the available dry pea stock in South Africa (233 MT); there was no impact on the local demand and supply of dry peas to the South African market. The evaluation team found that UMCOR’s data for three distribution dates (with one date prior, one on the day of distribution, and one the week following) showed no change in price on any products; however the limited time series means this finding is also inconclusive. Two of fourteen vendors interviewed during the evaluation team’s field visit reported LRP distributions as leading to a fall in commodity prices. The evaluation team’s analyses of existing and available data indicate that this project had no likely impact on the prices of the affected commodities. Please see Annex 3 and the market impact chapter of the main body of the report for further detail.


Zimbabwe’s cereal supply/demand balance for the 2010-2011 marketing year shows an uncovered national cereal deficit of 111,000 MT. Land reform efforts that began in 2000 continue to impact agricultural production, and over the years low production has necessitated average yearly imports of 500,000 MT plus an additional 170,000 MT in international assistance. A Zimbabwean ban on GMO foods has meant that imports must be certified GMO-free before entering the country, for both commercial trade and international assistance. The LRP field project for Zimbabwe complied with this testing regulation for importing maize from South Africa, though it was not an impediment to delivery. UMCOR reported that the Zimbabwean government re-introduced a 10 percent duty on imported maize meal products and a 15 percent duty on imported cooking oil during the project lifecycle.



According to FEWSNET, the re-imposition of the duty on maize meal and cooking oil in August may have exerted upward pressure on prices, although it was too early to gauge the impact of this policy on producers. UMCOR also reported a Zimbabwe Electricity Regulatory Commission tariff hike: tariffs were increased from $0.153 US cents per kilowatt hour to $0.983. Effects of this price rise were felt in the project district of Chipinge, where millers use electricity to power milling machinery. As UMCOR pointed out in its report to USDA, both of these interventions – the import taxes and electricity tariffs – took place in the last month of the project. They are therefore unlikely to have impacted the market during the project implementation period. With regard to the South African market, UMCOR reported that no government intervention took place at the time of the food procurement. USAID recently funded a $45 million, two-year Title II program, implemented by a consortium made up of CRS, ACDI/VOCA, and Care International. USAID reported that in FY 2011 this project was projected to distribute in-kind food aid totaling 7,590 MT. Approximately $1,000,000 was provided through the Bill Emerson Humanitarian Trust (BEHT) for Zimbabwe in 2010. An additional 22,010 MT of food aid was provided through an emergency Title II program with WFP in 2011. WFP procured regionally for food aid for Zimbabwe with funding of $5,717,900 through USAID’s Emergency Food Security Program (EFSP). Through interviews with the local WFP office, the evaluation team also learned of WFP’s efforts to procure locally when possible, but that these efforts were hampered by low production levels and high costs of purchasing from smallholder farmers and traders in Zimbabwe.


Hard tenders were used to buy 1,291 MT of maize, 233 MT of pulses (yellow peas), and 89 MT of vegetable oil in the regional market of South Africa.


For three procurements, the average overall time was 70 days for procurement, of which 58 days elapsed during contracting and 12 days during delivery. These figures were materially the same for all three commodities.


The average costs per MT for procurements were as follows: • Maize cost $285.00 per MT plus $152.85 per MT for TSH; • Yellow peas cost $508.00 per MT plus $120.22 per MT for TSH; and • Vegetable oil cost $2,097.00 per MT plus $151.23 per MT for TSH.

ii. Assess:


UMCOR’s reporting was found to be compliant with requirements.



UMCOR reported no measurable impact on producers in the procurement market in South Africa as a result of the LRP procurement. The procurement volume made up 0.188 percent of the white maize yield for Gauteng Province (where the maize was sourced). Similarly, UMCOR procured 89 MT of cooking oil from South Africa, which was just 0.037 percent of South Africa’s total cooking oil production. UMCOR purchased a mere 1.2 percent of the available dry pea stock in South Africa. Vendors interviewed by the evaluation team called these procurements “very small” relative to their other contracts.


Reporting Requirement Notes bb Markets

UMCOR reported that there were no measurable impacts in the procurement market, or on distribution markets as a result of their procurements and distributions. The evaluation team found it unlikely that the LRP procurements of the listed commodities in South Africa had any discernible or meaningful impact on the prices of these commodities in South Africa (the source market). Data for three distribution dates (with one data point prior, one at distribution, and one following) provided by the project for distribution zones show no change in price on any products. The evaluation team’s analyses of existing and available data indicate that this project had no likely impact on the prices of the affected commodities. Please see Annex 3 and the market impact chapter of the main body of the report for further detail.

cc Low-income consumers Given the findings of no likely impact on consumer prices, the evaluation team estimates no impact on non-recipient low-income consumers.

dd Program recipients Evaluation team interviews with LRP recipients indicated that program recipients have reduced spending on maize and reduced levels of piecework performed for others as a result of receiving food. Other coping mechanisms that were reportedly reduced included reliance on friends or relatives for food, buying food on credit, sending household members elsewhere to be fed, and collecting wild food or hunting. Over 60,000 vulnerable households and over 7,500 chronically food insecure households were supported during the lean season.

III Time of delivery For UMCOR’s three procurements, the average overall time was 70 days for procurement, of which 58 days elapsed during contracting and 12 days for delivery. These figures were materially the same for all three commodities.

iii. Compare:

IV Quality and safety assurances No problems were found regarding food quality and safety with these regional procurements. Testing was carried out by the private company SOCOTEC, and found no aflatoxin or other contaminants or safety or quality concerns in the commodities imported from South Africa. One exception was the high proportion of broken or damaged yellow peas, due to low moisture content, but this was not related to safety specifications. Additional testing carried out by the Department of Research and Special Services, Ministry of Agriculture, found the maize non-GMO and free of infestation as well.


FIGURE 1: MAIZE PRODUCTION, ZIMBABWE, 2000-2011, IN THOUSANDS OF MT

Source: Crop and Food Security Assessment Mission WFP/FAO Reports Various years, Agritex (Department of Agricultural Technical and Extension Services) and Central Statistics Office/Zimstats FIGURE 2: WHOLESALE MAIZE PRICE MOVEMENTS, SOUTH AFRICA, 2000-2011, IN

RAND/MT

Source: GIEWS, FAO

0

200

400

600

800

1000

1200

1400

1600

1800

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011

Prod

‘000

MT


ANNEX III. IMPACT ON INDIVIDUAL MARKETS

This section presents an analysis of market impacts for each project, related to and supporting the section on market impact in the main body of the text. For each analysis, the evaluation team chose a combination of methods depending on and best suited to the available data. In doing so, the team was guided by the following sequence: 1) examining available time series data, 2) using the price elasticity of supply to elicit price movements based on market throughput and procurement quantities, 3) using analyses available in project final reports, which sometimes included work by Cornell University in the framework of the Learning Alliance, and 4) going back to quantitative and qualitative data collected during field visits by the evaluation team in 2011. The combinations used in each analysis reflect the project characteristics, the availability and quality of data and of other analyses, and whether or not the evaluation team had visited the project in the field.


1. BANGLADESH Land O’Lakes implemented the LRP pilot project in Bangladesh, where it distributed locally produced cereal bars between the end of December 2010 and the end of the project in late September 2011. Table 1.1 shows the primary commodities used in making these bars. For the purpose of this analysis, the evaluation assumes that peanuts and sesame seeds are cash crops, of which almost all production enters the market.

TABLE 1.1: COMMODITIES, LOCAL PRODUCTION, AND QUANTITY USED 72

Commodities Local Production (MT) Purchased to Produce Cereal Bars (MT)

LRP Purchases as Percent of Total Production

Peanuts 40,000 211 0.53%

Sesame Seed 40,000 53 0.13%

Chick Peas 10,000 73 0.73%

Rice73 32,000,000 40 0.0001%

Source: Bangladesh Local and Regional Procurement Project: Final Report on Project Requirements, Land of Lakes Final LRP Project Report The evaluation of market impacts of LRP on Bangladeshi markets draws on wholesale and retail prices collected around the time of the procurements by Land O’Lakes. These data sets are limited to a two-week period around each purchase date, totaling three data points per procurement. The evaluation team summarized one such database and summarized trends in a sample take from another, and assessed variation in average monthly rice prices during specific procurement months.

72 The report gave no indication of the period over which this total applies, although the magnitude suggests that these data represent total production over either the 6-month procurement period or a complete year. 73 This item was listed as “Rice” in the Land O’Lakes report. Since it is actually “puffed rice” that is a cereal bar ingredient, that is the likely meaning here. This is true for at least two instances in the report, and in this market analysis; rice is listed when it is likely that puffed rice is the commodity monitored and purchased.


TABLE 1.2: LOCAL AVERAGE MARKET PRICES (RETAIL & WHOLESALE) IN CAPITAL CITY MARKETS (USD/MT)

Date of Procurement

Ingredient Commodities

Local Average Market Price/ MTs (three days

before procurement)

Local Average Market Price/ MTs (one day after procurement)

Local Average Market Price/ MTs (two days

after procurement)

Retail Wholesale Retail Wholesale Retail Wholesale

11/24/2010 Puffed Rice 794 712 799 724 789 728

Peanuts 1,201 1,148 1,155 1,124 1,196 1,151

Chick Peas 819 796 824 803 824 806

Sesame Seeds 828 1,090 811 1,035 811 1,068

3/30/2011 Puffed Rice 735 651 735 652 734 655

Peanuts 1,257 1,076 1,258 1,076 1,261 1,073

Chick Peas 938 876 942 877 942 874

Sesame Seeds 943 1,086 948 1,081 949 1,086 Source: Adapted from Land O’Lakes’ Final LRP Report, 30 September 2011

Note: The prices obtained from market vendors were per kilogram and have been converted to per MT prices. It is likely that prices per MT were lower, as vendors lower prices for bulk purchases. Retail prices for sesame seeds in markets in Dhaka were not available, so the retail prices listed for sesame are averages for all markets monitored in Bangladesh.

A second Land O’Lakes database provided similar data for 12 procurements for each of the commodities covered in Table 1.2. The worksheets provided a range of tabular data for wholesale and retail prices but the evaluation team judged that re-tabulation so as to aggregate the data would not be efficient, particularly as the data underpinning Table 1.2 appear to be a representative subset of this latter database. The team examined a sample of the price data spanning the range of procurement dates. For puffed rice, prices remained unchanged for all procurements. For chickpeas, prices remained unchanged for nine of eleven procurements, with others varying up or down by up to 4 percent. For peanuts, prices remained unchanged after seven of ten procurements, varying up or down by 1 percent (less than 2 %) in the other cases. For sesame, prices remained unchanged for nine of the ten procurements, varying up or down by 1 percent in the other cases. There is thus no systematic evidence of price changes due to local procurements of these project inputs.

Land O’Lakes also made available average monthly wholesale rice prices for Bangladesh (not shown). These data omit the price for November 2011, a rice procurement month. From March 2011 (during which one procurement took place) to April 2011 prices fell by 3 percent. From July 2011 to August 2011 (during which two procurements took place) prices remained constant. A significant price rise would not be expected when the tonnage purchased over eight months equaled 0.0001 percent of national annual production. Indeed, these aggregate data suggest falling or stable prices from the purchase month to the following month. This additional information corroborates the conclusion that market impact was unlikely for all four commodities considered.


2. BURKINA FASO Catholic Relief Services (CRS) implemented an LRP pilot project in Burkina Faso. The project procured 157.10 MT of cowpeas, 628.10 MT of millet, and 72.24 MT of vegetable oil. All commodities were purchased and distributed in Burkina Faso.

The evaluation team used wholesale market prices in procurement markets and a Cornell University assessment as the basis of its evaluation of LRP-related market impacts in Burkina Faso.

CRS estimated that in terms of available supply in the source markets, the overall quantities of millet and vegetable oil procurements represented less than 1 percent, while the total quantity of cowpea procurements accounted for 3 percent. Thus, the purchase of cowpeas was the only significant procurement in terms of commodity availability.

The time series price data available for procurements consisted of three data points – one week before, the day of, and one week after procurement. Table 2.1 shows these procurement-market prices for the three commodities and the three procurement periods. Millet prices rose 43 percent in the week before the procurement and by 25 percent in the week following the procurement. For cowpeas, prices rose 2 percent in the week preceding the procurement and by 11 percent the week after. Vegetable oil prices remained static across both these periods. The procurements took place at the end of Burkina Faso’s cool dry season, after small farmers had sold the majority of the production that they considered to be surplus to pay for household consumption needs. The procurement period is thus typically a season of rising prices, but not to the degree that the data shown in Table 2.1 reflect. The assertion that LRP may have played a role in provoking these price rises cannot be discounted.

TABLE 2.1: MARKET PRICE DATA FOR LRP COMMODITIES

IN PROCUREMENT MARKETS

Commodity Procurement Dates in 2011

Wholesale Market Price (USD/kg)

1 Week Before

Procurement Day

1 Week After

Millet Feb 17 - Mar 07 0.28 0.40 0.50

Cowpeas Mar 17 - Apr 07 0.61 0.62 0.69

Vegetable Oil Feb-17 2.60 2.60 2.60 Source: Upton, Joanna. Local Food for Education: LEAP Burkina Faso LRP Final Project Report.

Separately, Cornell University performed an analysis of market impact using a different database of millet and cowpea prices over 10 years in seven Burkinabè markets in the project’s procurement and distribution zones. The small procurement volumes prompted the analysts to combine them with similar purchases by WFP and estimate the joint effects of these procurements on market prices. These combined procurements impacted retail prices of the two commodities in procurement and distribution zones with statistically significant impacts for millet in the distribution markets one month and two months after the distributions, and for cowpea purchases in procurement zone prices one month after the procurement. However, this result does not provide evidence that can be ascribed uniquely to the LRP project.

In conclusion, there is mixed evidence about the influence of the procurements made under the CRS Burkina Faso LRP project on market prices. Market impact is possible.


3. CAMBODIA The International Relief and Development (IRD) LRP pilot project in Cambodia procured 36.1 MT of canned fish, 3.95 MT of iron-fortified fish sauce (IFFS), 114 MT of rice, and 11.8 MT of vegetable oil. Beneficiaries also received the following commodities through a voucher program: rice (225.1 MT), canned fish (3.8 MT), instant noodles (7.5 MT), and vegetable oil (4.4 MT). All commodities were purchased in Cambodia. Rice and IFFS were produced in Cambodia, and canned fish and vegetable oil were produced in Vietnam. In all cases, IRD distributed the commodities in Cambodia.

The timeline of the procurement of commodities is shown in the following table.

TABLE 3.1: COMMODITY PROCUREMENT DATES, IRD CAMBODIA LRP PROJECT

Procurement Approach

Quantity (MT) Commodity

Date of Issuance of Tender, or Equivalent

Date of Signature of Contract, or Equivalent

Date of Arrival of Commodities or First Voucher Purchases

Vouchers 3.76 Canned Fish 1/13/2011 2/17/2011 3/2/2011 Hard Tender 12.50 Canned Fish 1/1/2011 2/1/2011 2/2/2011 Hard Tender 23.60 Canned Fish 3/9/2011 3/31/2011 4/6/2011 Direct Purchase 1.95 IFFS 12/28/2010 2/23/2011 3/11/2011 Direct Purchase 2.00 IFFS 4/22/2011 5/11/2011 5/23/2011 Vouchers 7.47 Instant Noodles 1/13/2011 2/17/2011 3/2/2011 Hard Tender 41.50 Rice 12/28/2010 1/19/2011 1/31/2011 Hard Tender 72.50 Rice 3/10/2011 3/31/2011 4/6/2011 Vouchers 225.11 Rice 1/13/2011 2/17/2011 3/2/2011 Vouchers 4.38 Vegetable Oil 1/13/2011 2/17/2011 3/2/2011 Hard Tender 4.50 Vegetable Oil 1/1/2011 1/26/2011 1/31/2011 Hard Tender 7.27 Vegetable Oil 3/9/2011 3/31/2011 4/2/2011

Source: Adapted from discussions with IRD

The evaluation team assessed the presence of market impacts due to procurements of the above commodities by analyzing market price series data, then by using the short-run price elasticity of demand.

For each of the procurements, there are three time series data points, drawn from the commodity source markets. These data points are the retail market prices one week before the procurement, the day of the procurement, and one week after the procurement, except for rice where only secondary data on wholesale prices was available. This narrow range limits the analysis of potential market impacts due to procurement, but they do give an indication of any short-term price impact of LRP purchases. Graphs of the retail market prices of the four LRP commodities that were procured through hard tenders and direct purchases74 are presented in Figures 3.1 to 3.4.

74 Comparable price data were not available for voucher procurements


FIGURE 3.1: RETAIL PRICE OF CANNED FISH IN PHNOM PENH

Source: Adapted from USDA Local and Regional Procurement Pilot Project, Evaluation Report –Final LRP Project Report. International Relief and Development/Cambodia

FIGURE 3.2: RETAIL PRICE OF VEGETABLE OIL IN PHNOM PENH


2.05

2.1

2.15

2.2

2.25

2.3

2.35

2.4

2.45

26-Jan-11 1-Feb-11 8-Feb-11 ******** ******* 24-Mar-11 31-Mar-11 7-Apr-11

Pric

e pe

r Kg

(USD

)

Date

Retail Price in Phnom Penh of Canned Fish Around Times of LRP Procurements

1.44

1.45

1.46

1.47

1.48

1.49

1.50

1.51

19-Jan-11 26-Jan-11 2-Feb-11 ******* ******* 24-Mar-11 31-Mar-11 7-Apr-11

Pric

e pe

r Kg

(USD

)

Date

Retail Price in Phnom Penh of Vegetable Oil Around The Times of LRP Procurements


FIGURE 3.3: RETAIL PRICE OF IFFS IN PHNOM PENH


FIGURE 3.4: WHOLESALE PRICE OF LOW GRADE RICE IN PHNOM PENH


2.70

2.75

2.80

2.85

2.90

2.95

3.00

3.05

16-Feb-11 23-Feb-11 7-Mar-11 ******** ******** 4-May-11 11-May-11 18-May-11

Pric

e pe

r Kg

(USD

)

Date

Retail Price in Phnom Penh of Fish Sauce Around The Time of LRP Procurements

0.38

0.39

0.4

0.41

0.42

0.43

0.44

0.45

12-Jan-11 19-Jan-11 26-Jan-11 ******** ******** 24-Mar-11 31-Mar-11 7-Apr-11

Pric

e pe

r Kg

(USD

)

Date

Wholesale Price of (Low Grade) Rice in Phnom Penh Around The Time of the LRP Procurements


Of the four commodities, only the first procurement of canned fish and the second procurement of vegetable oil suggested the possibility of price spikes within one week of the procurement. In the case of the canned fish in February 2011, the price rose within one week after the procurement by $0.22 on a base price of $2.20, or 10 percent. Assuming a short-term price elasticity of +0.50, and with a change in price of +0.10 (i.e. a 10 percent increase), the corresponding change in supply would have to have been in excess of 5 percent. In other words, the procurement would have to have added at least 5 percent to the total Phnom Penh market throughput to have created a price spike of this magnitude. IRD bought half of all its canned fish, or 16.26 MT, in February, so the market throughput in that week would have to have been less than 325.2 MT in the same time frame for this price rise to have been a possible consequence of LRP. This evaluation did not have access to data on the Phnom Penh market throughput of canned fish, and so did not pursue this line of analysis. Similarly, the evaluation team did not have market throughput details for vegetable oil. If the entire increase in its price in March 2011 was attributed to LRP, then a price rise of $0.01 on a base price of $1.46, i.e. an increase of around 0.7 percent, would have to be explained. Given a short-term elasticity of supply of +0.40 for cooking oil, the change in quantity bought that week due to LRP would have been 0.28 percent. Since IRD bought 7.27 MT of vegetable oil in March, this would mean that the inhabitants of Phnom Penh would have to be consuming 2,596 MT of vegetable oil weekly, which seems unlikely for a city of a population of approximately two million.

The evaluation team concluded that LRP may have had an impact on the price of canned fish, although the information needed to test this conjecture was not available. Based on the minimal time series data available for this analysis, there is no indication of any other price impacts due to LRP.


4. CAMEROON The LRP project in Cameroon was implemented by WFP, who used hard tenders to procure 543 MT of beans, 1,790 MT of maize, and 329 MT of sorghum.

To assess the market impact of local procurements, the evaluation team used analysis based on the short-run price elasticity of supply and an analysis of monthly price data for sorghum, maize, and beans.

The team assumed that Cameroon farmers market 20 percent of their cereal production, i.e. 124,400 MT of sorghum, 34,600 MT of maize and 16,800 MT of beans; and applied the following short-run price elasticities of supply: sorghum +0.16, maize +0.23, and beans +0.25. Inserting these values along with the total procurement volumes into the equation for the price elasticity of supply yielded corresponding percentage price rises of: 1.65 percent (sorghum), 22.5 percent (maize), and 12.9 percent (beans).

The trends in monthly average wholesale prices in the procurement regions for each commodity for + 3 months around the delivery dates, using data made available by WFP, are shown below:

• Sorghum: Dec 2010 - June 2011 (Figure 4.1) • Maize: Nov 2010 - July 2011 (Figure 4.2) • Beans: Oct 2010 - Aug 2011 (Figure 4.3)

FIGURE 4.1: RETAIL PRICE OF SORGHUM IN THE FAR NORTH REGION OF CAMEROON

Source: Adapted from USDA Local and Regional Procurement Pilot Project, Evaluation Report –Final LRP Project Report World Food Programme/Cameroon


FIGURE 4.2: RETAIL PRICE OF MAIZE IN THE NORTH REGION OF CAMEROON


FIGURE 4.3: RETAIL PRICE OF BEANS IN THE WEST REGION OF CAMEROON


WFP’s project final report provided some important information for the interpretation of Figures 4.1 – 4.3. Suppliers to WFP made the purchases for the pilot project in mid-November when they signed their contracts with WFP, and continued to purchase to meet the contracted amounts through February and March, i.e. over approximately three and a half months. Thus, purchases took place as processes, rather than as one or two events. However, the profile of purchases over time is unknown, so there were no likely single price spikes for any of these commodities, purchased in smaller quantities over time.

Figure 4.1 shows that the sorghum price fell initially during the November – March period and then rose again. The evaluation team surmised that the bulk of the purchases took place in February and March, at the very end of the purchase period, which would explain the price rise of about 4 percent over the January –


March period. Without more information about the time profile of the purchases, the evidence is inconclusive.

Figure 4.2 shows that over the maize procurement period, maize prices fell by about 5 percent from November to December and then rose by about 30 percent over the next two months, comparable to the 23 percent price rise posited by the elasticity estimate. However, it is impossible to explain the continuing price rise at about the same rate until April, after the end of the maize procurement. Further, WFP staff reported no impact of their procurement on the quantity or price of maize, despite additional demand from Chadian and Nigerian traders and from buyers from Maiscam, the sole factory in Cameroon that produces maize meal.

Figure 4.3 shows a 33 percent bean price drop from November to December, followed by a 43 percent price rise for beans from December to January. The “yo-yo” effect of successive inter-monthly price rises and falls of 25 to 30 percent continued until March. This pattern could be consistent with two waves of bean purchases in November and January, but the evaluation team lacked details of the suppliers’ purchasing calendars to confirm this. Furthermore, if there were two waves of bean procurement, the price rise that elasticity calculations project would have to be halved, from a 12.9 percent increase to two 6.4 percent increases, significantly less than the rises discussed above. WFP’s report largely attributed the November price spike to extra demand from merchants in Congo, Equatorial Guinea, and Gabon in preparation for the Christmas festivities but noted that WFP demand would also have played some part: “[T]he perceived impact of WFP’s procurement on the beans market is considered to be limited, given that the price of beans decreased on the following month when the contracted suppliers were continuing their purchases in local markets.”

In conclusion, projected price rises derived from elasticity estimates for beans, and particularly maize, of the order of 10 and 20 percent respectively, suggest significant market disruption due to procurements of these staples in the different Cameroon production zones; similar estimates for the price rise due to procurements of sorghum of about 2 percent suggest a minor impact. However, price evidence over the purchasing period and the observations of WFP Cameroon staff in the field suggested that these estimates cannot be confirmed without more information about the profile of purchases by WFP suppliers over a three and a half month period. At the moment, no clear evidence suggests market impacts due to LRP for any of these three commodities.


5. CHAD WFP implemented an LRP project in Chad, in which it purchased 1,512 MT of maize in Cameroon, and 1,111 MT of cowpeas in Niger. Contract dates for maize and beans were, respectively, March 2, 2011, and March 21, 2011.

The evaluation team analyzed monthly price data made available by WFP to assess the likely market impacts of this project’s purchases of cowpeas and maize in their procurement zones.

Figure 5.1 shows monthly wholesale cowpea prices in Maradi, Niger, the procurement location. Over the next month the mean price fell by 39 percent per MT. The evidence does not suggest that the additional demand due to the procurement brought about a price increase.

FIGURE 5.1: COWPEA PRICES IN THE MARADI REGION OF NIGER (CFA FRANCS PER MT)

Source: Adapted from USDA Local and Regional Procurement Pilot Project, Evaluation Report Emergency Food Assistance to Drought-Affected Populations in Chad: WFP Chad Final LRP Project Report


Figure 5.2 shows mean monthly retail prices for maize in Garoua in Cameroon’s Northern Region, where procurement took place. From March (the procurement month) to April 2011, prices rose 6 percent from $29.20 to $31.00 per 100kg. The LRP project may have played a role in this price increase, but figure 5.2 clearly shows that from December 2010 to May 2011 maize prices increased monotonically; i.e. this particular price increase forms an integral part of an upward seasonal price trend underway before the procurement and continuing after it. A price impact attributable to the LRP project is therefore unlikely.

FIGURE 5.2: MAIZE PRICES IN GAROUA, CAMEROON ($/100 KG) JULY 2010 – JUNE 2011

Source: Adapted from USDA Local and Regional Procurement Pilot Project, Evaluation Report Emergency Food Assistance to Drought-Affected Populations in Chad: WFP Chad Final LRP Project Report

In summary, the evaluation team found no compelling evidence of an impact on the market prices of maize or cowpeas caused by the procurements made under this project.


6. REPUBLIC OF CONGO WFP implemented the LRP project in the Republic of Congo to support refugees from the Democratic Republic of Congo (DRC). WFP purchased 1,458 MT of “local rice” from two suppliers based in Bumba, DRC and distributed it in the Republic of Congo’s Likouala Province.

WFP’s final project report notes that the refugees’ rice distribution was a small fraction of production and that their demand was highly elastic. It estimated WFP’s procurement at approximately 1.5 percent of total production (1,458 MT of an estimated 99,000 MT).

The evaluation team assessed the potential for this regional procurement’s impact on prices based on the short-term price elasticity of supply, price trends at key procurement moments in the market where the rice shipments were assembled and dispatched, price trends in monthly price data at a key regional market, and WFP’s focus group results.

A price increase of 5.9 percent would be necessary to be able to attribute the rise to additional LRP demand for rice, based on a short-term price elasticity of supply for rice of +0.25 and the purchase and production figures given above.

Figure 6.1 shows rice prices in the Bumba market for locally produced rice. Note that the graph shows trios of prices pertaining to each of the three rice shipments75 from Bumba to the Republic of Congo distribution zone: one week before, on the shipment date, and one week afterward. Lacking the dates on which the two Bumba rice suppliers procured their rice, WFP used the dates of the shipments instead.

FIGURE 6.1: RETAIL PRICE, BUMBA (DRC) MARKET, LOCALLY PRODUCED RICE

Source: Adapted from Fossi, Filipo. USDA Local and Regional Procurement Pilot Project, Evaluation Report–Final LRP Project Report World Food Programme/Republic of Congo

The harvest period for rice in Bumba is mid-July to late August, explaining the drop in prices from mid-July to mid-August, which the WFP report attributes to rice farmers coming to Bumba market to sell their

75 The evaluation team used contract signature dates, or their equivalent, as reference dates to search for procurement price impacts. In cases where contract data are not available, as in the referenced sentence above, the team uses the dated in the commodity procurement cycle that are available.

0

100

200

300

400

500

600

700

One weekprior

shipment(baseline)

1st shipt:July 13th

One weekafter shipt

One weekprior shipt(baseline)

2nd shipAugust 18th

One weekafter shipt

One weekprior shipt(baseline)

3rd shipt:Sept. 12th

One weekafter shipt

Reta

il Pr

ice,

Con

gole

se F

ranc

(CDF

)


surpluses. Note in Figure 6.1 that the price of this rice was completely stable from August 18 (the date of the second rice shipment) to September 12 (the date of the last shipment). The price information in Figure 6.1 is from the key supply market but it provided little context for the evaluation team to make firm inferences about any impact of the LRP Project. For instance, whether the stable prices from August 18 to September 12 are due to the absence of an LRP effect or prices would have fallen for some other reason without the compensating effect of the additional LRP demand cannot be said. The minimum that can be concluded is that there is no direct evidence from the data underlying Figure 6.1 of a price rise that is unambiguously attributable to the LRP project.

Figure 6.2 shows the retail price of rice in Mbandaka, a city of about a million inhabitants and the major rice consumption center between Bumba (DRC production zone) and the distribution sites in Likouala Province in ROC, for calendar years 2009 and 2010, as well as the first seven months of 2011. WFP estimates that Mbandaka consumes 7 percent of the Bumba region’s rice production and, though it does not provide information about Mbandaka’s other sources of rice, WFP assumes that the price of rice in Mbandaka shows some similarity to that in Bumba. It is the rice market closest to Bumba for which times series of monthly prices are available.

Two conclusions can be drawn from the data displayed in Figure 6.2. First, data for 2009 and 2010, years without an LRP project, show a seasonal price fall from May or June through to a minimum in August –October. The available 2011 data do not cover the August - October period; a fall in prices from June to July is consistent with the same period in 2010. Secondly, the contract was signed with the suppliers on May 30, 2011, so purchases to meet this contract in and around Bumba would have taken place in the June – August period to allow the shipments to take place on July 13, August 18, and September 12. From May to June, the price rose by 10 percent but fell again from June to July by 6 percent. A sequence of the May-June rise and June-July falls also occurred more emphatically, in absolute and percentage terms, in 2009, a year without LRP-related purchases. So it cannot be said that the LRP project generated price changes outside the historical seasonal norms for which data are available. Further, in analyzing Figure 6.2, it should be noted that the two suppliers may have made LRP project shipments from pre-existing stocks. If so, this would further minimize the likelihood of any market impact related to LRP project purchases.

FIGURE 6.2: RETAIL PRICE OF LOCALLY PRODUCED RICE IN MBANDAKA, DRC

Source: Fossi, Filipo. USDA Local and Regional Procurement Pilot Project, Evaluation Report–Final LRP Project Report World Food Programme/Republic of Congo


Within the context of the evaluation team’s inability to extract definite conclusions from available rice price data about the market impact of the LRP project, WFP provides additional empirical information: “A focus group with the presence of small vendors and consumers confirmed that, despite rice being an expensive product in general, the seasonal pattern is usual.” WFP concludes that rice supply from the large procurement zone is “easily” able to absorb the extra LRP demand, especially because procurement took place after the harvest. Further, it concludes that spreading demand in Bumba over three smaller shipments, rather than one large shipment, reduced the potential for a price spike at any point in time. The evaluation team concurred and concluded that that LRP project-related purchases had no noticeable impact on the retail prices of locally produced rice in the Bumba market or in the upstream markets.


7. GUATEMALA CRS implemented the LRP project in Guatemala between October 2010 and September 2011, procuring 1,224 MT of white maize, 147 MT of black beans, and 160 MT of Incaparina, a dietary supplement rich in protein76.

The evaluation team’s appraisal of possible market impacts attributable to the LRP project rested on the consideration of price trends from CRS data and an econometric analysis done by Cornell University.

Cornell’s analysis noted price decreases lasting up to three months for white maize. This finding was statistically insignificant and ran counter to the expectation that prices would have increased. More generally, Cornell University reported:

“The largest estimated effects were on the retail prices of yellow maize, a substitute for the white maize the program procured; CRS’ maize [local procurement] actions were associated with an estimated -4.5 to -5.8 percent decrease in retail yellow maize prices at lags of up to three months, although these estimates are not statistically significantly different from zero. Point estimates of the effects of CRS’ [local procurement] activities on retail bean and Incaparina [sic]food prices are typically ± 1 percent or less.”

The evaluation team visited Guatemala and found that its maize market was fully integrated. Consequently, the national market benchmark price at “La Terminal” market is representative of most Guatemalan markets and an analysis of its white maize prices should reflect commercial conditions in procurement markets. Table 7.1 was created using data from the CRS report and shows the average prices of white maize from 2005 to 2011. Note that the data do not cover the entire period of the procurements. Procurement contracts were signed in October and November 2010 and in April, June, and July 2011.

While price changes in 2011 could not be determined, movements of mean price between October 2010 and December 2010 could be compared to those in the same time frame during the previous years. Table 7.1 shows that from 2005 to 2009, mean prices fell by 7 percent in October-November and by 2 percent in November-December. The pattern was the same in October-November 2010 (-2%) but from November to December prices rose by 3 percent. This suggests that this price increase is not a seasonal effect and may reflect, at least in part, the effects of this particular procurement, though the conclusion is based on minimal data.

TABLE 7.1: PRICE OF WHITE MAIZE AT THE NATIONAL MARKET, IN QUETZALES PER QUINTAL

Year Months

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2005 91.08 94.42 93.82 88.92 88.31 94.46 96.77 93.5 84.08 91.69 79.38 75

2006 77.54 78.83 92.38 89.78 84.92 95.85 98.23 99.69 94.83 87.5 88.08 90

2007 104.8 116.8 120.1 115.7 116.2 128.2 137.7 137.4 128.5 109.6 95.42 90

2008 98.77 104.1 110.5 109.7 112.1 127.6 126.9 123.9 132.8 129.1 120.3 121.2

2009 127.7 128.9 135.5 135 130.6 132.5 139 138.3 127.3 108.7 102.8 102.7

2010 118.3 115.5 123.6 123.4 119.1 119.1 131.1 132.1 136.8 117.2 114.9 118

2011 140.1 157 175 173.9

76 Incaparina is made with made with ingredients such as cottonseed, corn, and sorghum flours, and yeast.


Source: Adapted from USDA Local and Regional Procurement Pilot Project, Evaluation Report Final LRP Project Report Catholic Relief Services/ Guatemala

Table 7.2 shows prices of maize the week before, during, and after procurements. The price fluctuations across the five procurements have no coherent pattern. This and the above finding suggest that it is unlikely that the procurements of maize under this LRP pilot impacted prices.

TABLE 7.2: PRICES OF MAIZE ONE WEEK PRIOR TO, ON THE DAY OF, AND ONE

WEEK AFTER PURCHASE

Date of Purchase

National Market Price One Week Prior to Purchase

(Quetzales/quintals)

National Market Price on the Day of Purchase

(Quetzales /quintals)

National Market Price One Week after Purchase (Quetzales /quintals)

12-Oct-10 115 100 110

17-Nov-10 110 110 110

7-Apr-11 170 171 160

6-Jun-11 172 175 170

7-Jul-11 210 215 215 Source: USDA Local and Regional Procurement Pilot Project, Evaluation Report Final LRP Project Report Catholic Relief Services/ Guatemala

Table 7.3 shows monthly average national prices for black beans. Procurement contracts were signed in October 2010, November 2010, April 2011, and July 2011. From October to November 2010 prices rose by 1 percent, and from November to December 2010 they rose by 3 percent. There is no data allowing an assessment of price changes related to the April and July 2011 procurements.

The pattern of price movements is similar in October, November, and December 2010 to that in those months in 2007-2009, except in December 2010 when prices dropped instead of rising. Specifically, the mean price changes over 2007-2009 were 2.8 percent for October-November (compared to 1 percent in 2010) and -4.4 percent in November-December (compared to 3 percent in 2011). The evaluation team thus concluded that the November 2011 procurement may have affected prices.

TABLE 7.3: PRICE OF BEANS AT THE NATIONAL MARKET, IN QUETZALES PER QUINTAL

Year

Months

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2007 258.1 249.2 257.3 257 250.8 263.1 260 262.3 272.7 269.6 292.5 281.1

2008 272.3 281.5 281.4 313.5 372.9 448.8 492.2 500.8 460.2 479.5 487.4 432.7

2009 425.5 436.3 396.4 410 429.5 436.2 445.2 427.6 431.4 430 438.3 421.5

2010 410.6 402 382.3 371.5 358.6 368.1 360 351.8 337.1 382.3 386.2 398.8

2011 417.9 405 376.7 362.2 - - - - - - - -



Table 7.4 shows prices of beans the week before, during, and after procurements, with no apparent trend that might allow a conclusion of a price impact due to bean procurements under the LRP project. The likelihood that the purchases of beans had an impact on prices is low.

TABLE 7.4: PRICES OF BLACK BEANS ONE WEEK PRIOR TO, ON THE DAY OF,

AND ONE WEEK AFTER PURCHASE

Date of Purchase

National Market Price One Week Prior To Purchase

(Quetzales/Quintals)

National Market Price On The Day Of Purchase (Quetzales /Quintals)

National Market Price One Week After Purchase

(Quetzales /Quintals)

12-Oct-10 375 355 390

17-Nov-10 340 373 390

7-Apr-11 375 312 300

7-Jul-11 360 370 370 Source: USDA Local and Regional Procurement Pilot Project, Evaluation Report Final LRP Project Report Catholic Relief Services/ Guatemala

The CRS report provides annual average prices of the fortified flour product Incaparina, from which it is not possible to infer any price impact due to the LRP project. However, during the evaluation site visits, a senior manager of that company stated that the project procurements were such a small part of the overall production and stockpile of Incaparina that they did not affect its price.

The evaluation team therefore considered that LRP-induced market impacts on the commodities procured by the project in Guatemala were unlikely.


8. KENYA World Vision implemented the LRP pilot project in Kenya, where it purchased 1,114 MT of maize, 204 MT of beans, 245 MT of corn soy blend (CSB), 82 MT of vegetable oil, and 10 MT of salt. All commodities were procured in Nairobi, Kenya, and distributed in Moyale, Kenya.

The evaluation team used price data and econometric analysis done by Cornell University to assess the potential impact of LRP on the project’s commodities. The final report comments on the significant price volatility for both maize and beans during recent years. The global food price crisis of 2007-2008 foreshadowed dramatic increases in maize and bean prices through late 2009, after which bean prices maintained their new high level while maize prices dropped back to pre-crisis levels. Figure 8.1 illustrates this volatility.

FIGURE 8.1: WHOLESALE PRICES FOR MAIZE AND BEANS IN NAIROBI, KENYA 7/2006-11/2010

Source: USDA Local and Regional Procurement Pilot Project, Evaluation Report –Final LRP Project Report World Vision/Kenya PEAK (Project for Emergency Assistance in Kenya) Cornell University analyzed the price impact of LRP on behalf of World Vision and the Learning Alliance for this project using an econometric analysis. It found a 10 percent increase in the price of maize that might have been a result of the LRP project, but noted that other market disruptions may be behind these increases:

“(i) we find no statistically significant maize price impacts in the central markets through which these competitive tenders were made, (ii) we do not find any estimated upward maize price effect of WFP procurements during the same period, although these were more than 40 times greater in average magnitude than World Vision’s maize procurements under the LRP pilot program, and (iii) we find no estimated effect of [local procurement] on price volatility.”

Figure 8.2 shows clusters of price data for the period around each of six maize procurements that were made in Nairobi. World Vision collected wholesale prices one week before each procurement, on the date of the procurement, and one week afterward. The prices rose during the first three periods considered, fell and then


rose around 1 July 2011, and then fell during the last two periods. The trend thus revealed appeared to be seasonal rather than LRP-induced. Indeed, there was no systematic trend in anticipation of the procurements or following them during the timeline illustrated in Figure 8.2.

FIGURE 8.2: WHOLESALE PRICES FOR MAIZE IN NAIROBI, KENYA: 2011

Source: Adapted from USDA Local and Regional Procurement Pilot Project, Evaluation Report –Final LRP Project Report World Vision/Kenya PEAK (Project for Emergency Assistance in Kenya) Notes: dates are in day/month/year format

Figure 8.3 provides the same information for beans as does Figure 8.1 for maize. The seasonal price trend is again evident, though the price of beans did not rise monotonically until the peak in June. However, bean procurements do not seem to systemically cause prices to rise. Procurements do coincide with the price rises in four of six cases but seasonality more convincingly explains this bias. An argument of price increases in anticipation of procurement is not convincing even though prices rose in the week preceding procurement in five of six cases: there is no evidence to suggest that information made available in advance provoked traders’ expectations for increased bean demand.

USD

/MT


FIGURE 8.3: WHOLESALE PRICES FOR BEANS IN NAIROBI, KENYA: 2011

Source: Adapted from USDA Local and Regional Procurement Pilot Project, Evaluation Report –Final LRP Project Report World Vision/Kenya PEAK (Project for Emergency Assistance in Kenya) Notes: dates are in day/month/year format

With this limited information, the evaluation team determined that there is no conclusive evidence of market impacts of the LRP project.

USD

/MT


9. MALAWI Under the LRP project, WFP procured 873 MT of maize, 656 MT of beans, and 815 MT of CSB. Commodities were procured and distributed in Malawi. WFP concluded contracts with suppliers from May 2010 to March 2011.

The LRP grant in Malawi funded part of the Purchase for Progress (P4P) program. P4P promotes the strengthening of Farmers’ Organizations (FOs) to allow them to supply food to WFP’s humanitarian operations within the framework of WFP’s transition from providing food aid to providing food assistance.

The evaluation team’s analysis of market impacts rests on a comprehensive analysis performed by WFP’s country office as well as data obtained from key informants.

WFP collaborated with the Ministry of Agriculture and Food Security of Malawi to reinforce the capacity of its Agro-Economic Survey (AES), with the aim of guaranteeing the collection of commodity price data in FOs’ local markets, the regional market nearest to the FOs where AES was already collecting data, and data for other key national markets. The regional and national markets serve as control markets. This capacity building results in a set of mostly retail databases with monthly prices of maize, cowpeas and substitutes, and soy (an ingredient for CSB), to allow comparison of local market prices with those in regional and national markets. All data span periods on either side of the procurements and, on the more established markets.

WFP Malawi’s final report does not perform econometric analyses but uses the data to test whether any of the six criteria77 flag potentially abnormal commercial behaviors within and between markets. For example, WFP examined if the price level in the local market exceeded the recent five-year average for that month in the nearest market with five years of data. Analysis of potentially abnormal market conditions flagged by these criteria took place in the context of recent developments in the enabling environment, including a glut of maize in Southern Africa coinciding with minimum farm-gate prices.

Maize

WFP reported that the model used limited purchased quantities of maize to 200 MT per FO or 300 MT per district to avoid raising prices. Maize prices remained lower than the five-year average in all markets monitored; the tonnage that would have provoked noticeable price spikes would have been significantly higher than these levels. In most markets, retail prices were also more stable than the five-year average. In the one case where volatility exceeded the five-year average, it was due to two pronounced price decreases, relative to surrounding markets, rather than the price increases that additional LRP demand might create. In some local markets, prices changed faster than the five-year average between certain months, but the WFP report authors explain that their data emphasized a given price change relative to the low price levels prevailing in 2010, in comparison to earlier years when price levels had been higher. They found one case where prices jumped from about 27 Kwacha/MT (K/MT) to about 34 K/MT for two weeks at the end of 2010, thereafter returning to their former level, which they were able to link to a purchase made under the LRP project.

From correlation analysis, rather than regression analysis that would allow controlling for exogenous variables, the authors formed the impression that markets were less integrated than they had predicted. If that is indeed the case, any price fluctuations in regional markets were more difficult to attribute to LRP.

WFP also systematically discussed the effects of LRP contracts with local focus groups. WFP paid the legally required floor price, higher than the market rate, when purchasing maize from FOs, and it paid a large part of the price up-front, which many traders do not do, thus encouraging some members to declare that they planned to increase their maize production in future seasons. Generally, changing such output goals requires increased prices over several seasons. Further, FOs reported more applications for membership for the same reasons. However, the impact of the procurements on prices in local or other markets was not discussed.

77 The criteria are listed in terms of price levels, price changes and price volatility.


Cowpeas

AES had minimal historical data on cowpeas, leading to unsatisfactory intertemporal comparisons with pulse substitutes. This left WFP’s price-impact monitoring entirely reliant on correlation of purchases with monthly price changes. However, WFP obtained cowpeas through repeated soft tenders and was often unaware of the origin(s) of a given delivered shipment. Therefore, it did not know which market price profile to compare for a potential price spike. The only price spike the authors found was modest and occurred after the purchase of 76 MT of cowpeas through soft tender. However, prices subsided quickly thereafter.

CSB

Malawian households do not normally purchase CSB, so there are no appropriate domestic price series to stand as references. “To further complicate analysis, most CSB processors purchase raw inputs at the beginning of the marketing season and then wait for contracts to begin manufacturing CSB; it is unclear how CSB tonnages and prices directly impact retail maize markets, or the small consumer retail market that exists in urban areas of Malawi for CSB.” Therefore the WFP analysts tried to examine the price behavior of its principle ingredients, maize and soy. Maize is discussed above. Soy prices in three out of the four markets examined in WFP’s analysis were very stable although they increased in the fourth market. The analysis of CSB prices done by WFP generated no more evidence of price spikes due to the LRP project.

In light of this discussion of the assessment conducted by WFP Malawi, the evaluation team concluded that the likelihood of any market impacts due to local commodity procurements during the implementation of the pilot project is weak.


10. MALI There were two separate LRP projects in Mali, each of which is discussed in a separate subsection.

10. A. Catholic Relief Services (CRS) LRP Project in Mali

CRS implemented the LRP project in Mali and documented their project in three separate reports. Table 10.A.1 shows that relatively small quantities of these commodities were purchased and delivered over a three month period in late 2010 and early 2011.

TABLE 10.A.1: COMMODITY QUANTITIES AND PROCUREMENT78 DATES, CRS PROJECT IN MALI

Procurement Approach

Quantity (MT) Commodity

Date of Issuance of Tender, or Equivalent

Date of Signature of Contract, or Equivalent

Date of Arrival of Commodities or

First Voucher Purchases

Hard Tender 0.98 Cowpeas 12/8/2010 12/28/2010 2/10/2011

Vouchers 6.69 Cowpeas 11/19/2010 12/20/2010 1/8/2011

Vouchers 31.45 Millet 11/19/2010 12/20/2010 1/8/2011

Vouchers 2.14 Rice 11/19/2010 12/20/2010 1/8/2011

Hard Tender 4.87 Rice 12/8/2010 12/28/2010 2/10/2011 Source: MSI discussions with CRS

The evaluation team assessed the market impact of this project through a combination of the use of short-term price elasticity of supply to estimate likely price changes and assessments of monthly price data in various markets.

Malian consumption of millet in 2010/2011 was 1,175,000 MT, according to CRS’ final project report, all produced domestically. A Michigan State University staff paper written in 1999 asserts: “only about 15-20 percent of total grain production enters the market.” Assuming the proportion may have risen since then, the team used the higher percentage (235,000 MT annually; 58,750 MT quarterly) to estimate sales activity. The LRP project purchase of 31.45 MT equals 0.054 percent of this quarterly figure. Taking a short-term price elasticity of supply of +0.16 for millet the team found a projected price increase of 0.33 percent from this additional demand.

Figure 10.A.1 shows monthly millet price fluctuations in major Malian markets at the wholesale level from October 2010 to March 2011. The trends suggest a relatively low level of spatial market integration across the country. The fluctuations in all markets exceeded 10 percent over this period, suggesting that the millet purchases financed under this project had no discernible impact on prices.

78 Some procurements were through vouchers, and therefore the dates correspond to the voucher program process outlined in the methodology section of the report.


FIGURE 10.A.1: MALIAN WHOLESALE MILLET PRICES

Source: Adapted from USDA Local and Regional Procurement Pilot Project, Evaluation Report –Final LRP Project Report CRS/Mali

The local procurements are more likely to have a market impact in local markets. Figure 10.A.2 shows monthly retail millet prices for markets local to the CRS project zone. In contrast to the profiles of Figure 10.A.1, Figure 10.A.2 shows trends of increasing millet prices over the procurement period, which correspond closely to the trend in Figure 10.A.1. The levels of millet prices vary according to the relative production of the agricultural zone in each market’s catchment area. The three markets shown in Figure 10.A.2 show prices at markets which are well connected by road to Mopti: Bankass (119 km), Douentza (186 km), and Djenné (115 km). Within the range of variation that these markets show, a 0.33 percent change would not be noticeable.

FIGURE 10.A.2: RETAIL MILLET PRICES IN LOCAL MARKETS



The LRP project purchase of rice was 7.01 MT against an annual market throughput of 1,634,000 MT, or a quarterly throughput of 408,500 MT. Using short-term price elasticity of +0.26 resulted in a projected price increase of 0.007 percent. This is even less likely to be noticeable than the market impact of millet. In addition, figure 10.A.3 shows that rice prices were falling over the project period so that, had the local procurement of rice had an impact, the impact would have had a stabilizing effect on local prices.

FIGURE 10.A.3: RETAIL RICE PRICES IN LOCAL MARKETS



With a national production of cowpeas in 2010 of 109,500 MT, assumed to be consumed within the country, the quarterly market throughput is 27,375 MT, so the projected price increase due to purchases of 7.67 MT would be 0.11 percent. Figure 10.A.4 shows that, over the course of the project, cowpea prices were increasing by between 3 and 14 percent monthly. Again the change due to local procurement would not be noticed.

FIGURE 10.A.4: RETAIL COWPEA PRICES IN LOCAL MARKETS


The net effect of this evidence was that the evaluation team concurred with CRS, who states in its market monitoring study:

“The conclusion … was that the LRP activity did not have any effect—positive or negative—on commodity prices in Mali during the period of the activity.” Indeed from the start, the small size of the LRP activity (a $106,000 pilot activity) negated the likelihood of any impact.


10. B. World Food Program (WFP) LRP Project in Mali

The WFP LRP project in Mali procured 2,251 MT of millet, 504 MT of sorghum, and 49 MT of cowpeas in fifty separate procurements between February 2010 and February 2011.

The evaluation team used detailed price information from WFP’s final project report to assess the likelihood of market impact.

The WFP report and associated Excel data file present data on the prices of millet and sorghum in the various markets one week before and one week after the project procurements.79 Table 10.B.1 reproduces these data, and shows that few of the market prices increased relative to the pre-procurement prices. The major exceptions are millet and sorghum on 10 February 2011 from the Ouré market and the wholesale prices collected from the proxy OMA Bougouni market (substituting for the actual procurement market). The mean price variation across all cases listed in table 10.B.1 equals 1.2 percent.

No additional price data are available for the OMA market to enable a more complete assessment of whether this price increase might have been due to the local procurement. However, table 10.B.1 shows that the purchase sizes of the two exceptions are 10.3 and 4.1 MT and that the mean procurement size is 70 MT. Larger procurements are more likely to raise prices than smaller ones, so these price spikes were not likely to have been caused by LRP when there was no parallel activity in the larger markets.

Again using data from WFP, the market throughput was 2,122 MT of millet and 3,165 MT of sorghum in the 2009-2010 season in the Sikasso region, where the procurement market of Ouré and the OMA proxy Bougouni market lie. Assuming that the 2010-2011 market throughput was more or less the same as that for the preceding year, the local procurements represent 0.5 percent of the regional market throughput of millet in that period, and 0.1 percent of the regional sorghum market throughput. These very small proportions further diminish the likelihood that the two price spikes around the times of the local procurements were the result of those procurements.

79 In some cases the prices at the market where these two commodities were purchased, prices were not monitored; in such cases, a nearby Observatoire du Marche Agricole (OMA) market where prices are regularly monitored by the Government of Mali was used as a proxy for the unmonitored market. No market data were available for cowpeas.


TABLE 10.B.1: WHOLESALE AND RETAIL PRICES OF MILLET AND SORGHUM IN MALI MARKETS BEFORE/AFTER PROCUREMENTS WERE MADE

Delivery Date Region Market

Location

OMA-Monitored

Market Cereal

Quantity Purchased

(MT)

P4P Price/MT

(FCFA)

Price – 1 Week Before

Delivery (FCFA)

Price – 1 Week After

Delivery (FCFA)

Price Variation (percent)

Price Type

20-Mar-10 Koulikoro Beleko Fana Millet 6.6 160000 120000 120000 0% Wholesale

5-Apr-10 Koulikoro Dioila Dioila Millet 37.8 155250 115000 112000 -3% Wholesale

8-Jan-11 Koulikoro Beleko Fana Millet 60 130250 110000 110000 0% Wholesale

10-Jan-11 Koulikoro Dioila Dioila Millet 50 130250 110000 110000 0% Wholesale

24-Feb-10 Sikasso Koutiala Koutiala Millet 2 140250 125000 125000 0% Wholesale

8-Feb-11 Sikasso Koutiala Koutiala Millet 50 145250 115000 115000 0% Wholesale

10-Feb-11 Sikasso Ouré Bougouni Millet 10.3 170250 100000 110000 10% Wholesale

25-Feb-10 Sikasso Sikasso Sikasso Millet 5.1 140250 175000 175000 0% Retail

9-Feb-11 Sikasso Sikasso Sikasso Millet 25 160250 175000 175000 0% Retail

9-Feb-11 Sikasso Sikasso Sikasso Millet 7.1 150250 175000 175000 0% Retail

18-Mar-10 Ségou Baraoueli Fana Millet 20 150250 120000 120000 0% Wholesale

18-Mar-10 Ségou Baraoueli Fana Millet 15.5 150250 120000 120000 0% Wholesale

23-Mar-10 Ségou Dioro Dioro Millet 15 147250 110000 105000 -5% Wholesale

9-Feb-11 Ségou Baraoueli Fana Millet 92 140250 110000 115000 5% Wholesale

9-Feb-11 Ségou Kono-bougou Fana Millet 57.7 140250 110000 115000 5% Wholesale

13-Feb-11 Ségou Kono-bougou Fana Millet 51.7 140250 110000 115000 5% Wholesale

16-Apr-10 Ségou Cinzana Ségou Millet 288 150250 150000 150000 0% Retail

26-Jan-11 Ségou Segou Ségou Millet 204 149500 125000 125000 0% Retail

12-Feb-11 Ségou Cinzana Ségou Millet 278.7 145250 125000 125000 0% Retail

3-Mar-10 Mopti Koro Bankass Millet 37.2 156920 140000 135000 -4% Wholesale

3-Mar-10 Mopti Koro Bankass Millet 23.8 156920 140000 135000 -4% Wholesale

4-Mar-10 Mopti Bankass Bankass Millet 28.7 156920 140000 135000 -4% Wholesale

19-Jan-11 Mopti Bankass Bankass Millet 30 140250 105000 110000 5% Wholesale


Delivery Date Region Market

Location

OMA-Monitored

Market Cereal

Quantity Purchased

(MT)

P4P Price/MT

(FCFA)

Price – 1 Week Before

Delivery (FCFA)

Price – 1 Week After

Delivery (FCFA)

Price Variation (percent)

Price Type

21-Jan-11 Mopti Koro Bankass Millet 70 135250 105000 110000 5% Wholesale

22-Jan-11 Mopti Koro Bankass Millet 100 135250 105000 110000 5% Wholesale

8-Jan-11 Koulikoro Beleko Fana Sorghum 150 130250 105000 110000 5% Wholesale

10-Jan-11 Koulikoro Dioila Dioila Sorghum 270 135250 110000 110000 0% Wholesale

10-Feb-11 Sikasso Ouré Bougouni Sorghum 4.1 140250 95000 105000 11% Wholesale

9-Feb-11 Sikasso Sikasso Sikasso Sorghum 50 135250 125000 125000 0% Retail

Source: Bishop, Helen and Pierre Traoré. Report on USDA-Funded Purchases 2009 – 2011, WFP Mali Final LRP Project Report


In their report, WFP offered the following possible explanation for these two apparent price spikes:

One explanation for this could be that in contrast to producers in other regions, Sikasso producers are better organized as a result of their involvement in cotton production and working with the CMDT.80 It is also worth mentioning that in February 2011 CMDT made timely payments for cotton. [T]here was no urgent need for farmers to sell dry cereal on the market. This, though unintentional, could possibly have created artificial scarcity, triggering price increase.

The conclusion is that the local purchases would have been unlikely to have caused detectable price rises. The overall conclusion reached by WFP in their report was: “Results indicate no significant price distortions occurring due to [LRP] purchases during the procurement period.” The evaluation team concurs.

80 CMDT is the Compagnie Malienne pour le Développement du Textile, the main cotton grower in the Sikasso region


11. MOZAMBIQUE WFP implemented an LRP project in Mozambique, where it distributed maize and cowpeas between April and July 2011. WFP’s suppliers were companies in Malawi, Mozambique and South Africa.

Two Malawi-based companies supplied 4,179 MT of maize: an initial maize tender resulted in a contract to supply 3,576 MT and a second maize tender resulted in a contract for 603 MT. A separate tender for cowpeas resulted in contracts for a Mozambican company to supply 136 MT and a South African company to supply 350 MT. All tenders specified non-GMO products, favoring countries with bans on GMO agriculture and limiting competition.

To conduct its analysis, the evaluation team used data collected from project stakeholders during field visits in Mozambique and Malawi in October 2011, as well as WFP Mozambique’s project final report.

Maize

WFP’s final project report discusses the market dynamics of cowpeas in Mozambique and of maize in Malawi and Mozambique. However, its analysis of market impact is limited to an assessment of one figure for white maize prices in Mutarara (in central Mozambique near the Malawian border) that contrasts the January – June 2011 price profile with the prices averaged over the same months for the previous five years. The report does not appear to take into account inflation over the period (See Figure 11.1). Without reference to any other plausible explanatory factors, it argues that maize distribution from mid-February to early March 2011 might have led to the price reduction observed during the same period. However, no claims are made about market impacts due to procurement.

FIGURE 11.1: RETAIL PRICES IN MUTARARA MARKET, MOZAMBIQUE

Source: Mutondo, Joao and Bruno de Araujo, Roger Bourgete, and Guy Mullin. “Market Assessment: Food Support for Protection and Promotion of Lives and Livelihoods of the Most Vulnerable People under USDA Grant. An assessment of USDA-funded WFP commodity purchase in Mozambique

Five factors appear to have contributed to the price trends observed for Malawian white maize in the first half of 2011. First, SENWES, the company that won WFP Mozambique’s first LRP tender for white maize, bought over 4,000 MT from January 26 until February 18, 2011. Much of the quantity needed for SENWES’ obligations under its WFP contract came from these purchases. On January 26, 2011, at the start of the purchases to satisfy the terms of its contract with WFP, SENWES was paying 28,000 Kwacha per metric ton


(K/MT), equivalent to $187/MT. Over the course of the 24-day purchase period culminating on February 18, 2011, prices rose 10 percent.81 Smaller traders sold the maize to SENWES: the ten largest suppliers provided 3,985 tons (with two providing over 1,000 MT each). SENWES staff assumed that their assertive buying explained the price increase over this 3.5 week period, which figure 11.2 illustrates.82

Second, this price rise was not entirely due to SENWES’ additional demand. At around the same time, four wholesale-trading competitors in Lilongwe were also buying to meet regional orders, and they appear to have cumulatively bought at least the same volume as SENWES to meet these orders. A 10 percent price rise from February 4 to March 1 may tentatively be attributed to this joint effect.

FIGURE 11.2: LILONGWE, MALAWI WHOLESALE WHITE MAIZE PRICES

Source: Mutondo, Joao and Bruno de Araujo, Roger Bourgete, and Guy Mullin. “Market Assessment: Food Support for Protection and Promotion of Lives and Livelihoods of the Most Vulnerable People under USDA Grant. An assessment of USDA-funded WFP commodity purchase in Mozambique Note: there were no trades in white maize at Agricultural Commodity Exchange for Africa (ACE), Lilongwe from 1st March to 26th May.

The Malawian maize market is distorted by government policy and by specific government actions. In February 2011, the Malawian government declared a new lower floor price for exports of maize for the new season starting at harvest in June.83 Once the big traders had met their regional orders, such as SENWES’ LRP order for WFP in Mozambique, they incorporated this new information into their market expectations, pushing prices downwards. This third factor contributed to the drop in maize prices after early March 2011. Separately, a fourth factor intervened on March 2, 2011: the National Strategic Reserve Agency (NSRA) announced the successful bidders for tranches of 30,000 MT of old maize it had proposed to sell off as part of a rotation of stocks; this placed further downward pressure on prices in the three month pre-harvest period that is traditionally a period of low cereal sales.84

The lower export floor price and the additional 30,000 MT on the market together explain the significant price drop over the three months until the start of the 2011 harvest, shown above. However, fifth, in contrast

81 The 28,000 – 29,000 K/MT price level was not historically high: over the last five years, prices in January have surpassed 60,000 K/MT. 82 The companies that won WFP tenders placed their bids through the Agricultural Commodity Exchange, which supplied the information for Figure 11.2. 83 The 2010 export floor price had been 35,000 K/MT. If the government does not grant export licenses (i.e. imposes an export ban) the demand for maize drops to the extent that the domestic price falls short of the export floor price. SENWES and its competitors were fortunate that WFP in Mozambique sought maize when the government was granting export licenses. 84 Some of this maize was five years old and therefore not of good quality.


to marketing years following poor harvests, the government did not impose an export ban following the 2010 harvest. This allowed the big traders to continue exporting as long as they could buy stocks from smaller traders.85

The price effect of the LRP project purchase of 3,576 MT is difficult to isolate but an estimation of orders of this magnitude is demonstrated below. Assuming the procurement accounted for 50 percent of additional maize purchases during February 2011, provoking a price rise of about 7 percent, the purchases used for the LRP project make up about a 3.5 percent increase. Importantly, SENWES did not make its January and February purchases with the knowledge that it would win the WFP LRP project tender in Mozambique. It was buying with that in mind but also with the possibility of resale to other parties. There is thus no direct causal link between this price rise and the LRP project purchases.

It is also important to note that, at the time of SENWES’ bulking to assemble the quantity of maize necessary for the WFP contract, Malawian farmers held almost no disposable stocks. Farmers generally sell all maize in excess of their subsistence needs within three months of the harvest, so by October, they scarcely respond to price incentives. By then, Malawi’s commercial maize stocks lay in the hands of traders and the NSRA. Thus the procurement in January and February could have had almost no effect on farmers. On the other hand, this procurement may have contributed to raising maize prices for consumers, including food-insecure consumers. However, in practice, the prices then dropped considerably due to government policy, which is a significantly more important determinant of maize prices in Malawi.

The Malawian government’s market information system tabulates weekly average retail prices for maize in markets around the country but with gaps, particularly during the first quarter of 2011.86 The figure below shows monthly average retail prices for three Lilongwe markets from November 2010 to June 2011, within the period LRP occurred. Visual inspection shows that there is little correlation between the price trends shown. It seems that the wholesale market at ACE focused on the export market and dealing with large tonnages that traders bought from other traders, and was largely segmented from the retail markets. This level of disparity is odd but does not appear to be an artifact of poor data: on the one hand, the ACE data came from carefully filed documentation on each sale, and on the other hand, two recent studies of the Malawian maize market based on government retail data make no mention of the low quality of those data.87 Therefore it is likely the machinations of procurements at the wholesale level had little, if any, impact on the retail market where most producers and food-insecure consumers would have sold and bought their maize.

85 An overvalued Kwacha and associated exchange controls lie at the heart of policy distortions in Malawi. The inability to export Kwacha-denominated profits at market value leads companies that would normally have nothing to do with the grain trade to export maize in order to be able to sell it outside Malawi for payment in hard currency. They are prepared to take a loss in order to avoid buying hard currency within Malawi at official rates. This puts downward pressure on maize prices. Separately, the difficulty that SENWES had in transporting the WFP maize to Beira stemmed from a lack of foreign exchange to buy the fuel for the trucks to be used. 86 In contrast, figure 11.3 features all wholesale trades at ACE during the period displayed. 87 See: Jayne, T., et al. (2010) Malawi’s maize marketing system. Prepared for the Agricultural Input Subsidy Programme, February 6; and Manda, E. (2010) Price instability in the maize market in Malawi, PhD thesis at the University of East Anglia, June.


FIGURE 11.3: RETAIL MAIZE PRICES IN THREE LILONGWE MARKETS

Source: Mutondo, Joao and Bruno de Araujo, Roger Bourgete, and Guy Mullin. “Market Assessment: Food Support for Protection and Promotion of Lives and Livelihoods of the Most Vulnerable People under USDA Grant. An assessment of USDA-funded WFP commodity purchase in Mozambique

Market impact information were not available for WFP’s second, 603 MT LRP maize procurement from Malawi.

Cowpeas

Annual Mozambican production of cowpeas averages 10,000 – 12,000 MT but poor rains reduced the 2011 harvest to 6,000 – 7,000 MT. As cowpeas are a cash crop, almost all of this production is marketed entirely within Mozambique. All of the 136 MT of cowpeas that Maviga, a large Mozambican trader based in Nampula Province in northern Mozambique, supplied to WFP came from Nampula province, a major production zone in northern Mozambique. Maviga makes 25 – 30 percent of its core purchases directly from farmers at harvest from April through June, after which it obtains the rest through a local buying agency that has 45 small subcontractors who have bought from farmers. Maviga also buys cowpeas from other large Mozambican traders. In total, it buys about 3,000 MT annually. To fulfill its LRP contract with WFP, Maviga used 46 MT of leftover stocks and purchased a further 90 MT from the local buying agency.

Thus, the quantity purchased in direct response to LRP needs was 90 MT of an assumed 2011 marketed surplus of 6,000 MT, or 1.5 percent. This seems to have had some small impact on the price of cowpeas, but the Maviga manager interviewed by the evaluation team argued that the company sells more or less the same quantity of cowpeas each year, and that the WFP contract did not have any immediate impact on supplies or prices. The purchases started before the April 2011 harvest and continued into the harvest period. So if the LRP project had caused price rises, these would have benefited both farmers and traders while penalizing consumers. However, the evaluation team did not gather detailed information on the split of profits along the market chain; it appears that any price increase was negligible. 2011 cowpea prices were 14 metical/kg in April, rising to 17 metical/kg in July. These prices were 17 – 33 percent higher than normal. Observers attributed this increase to the poor 2011 harvest, which appears to have been a much more important factor influencing price than the procurement.

Durban-based JLR is a relatively large company integrated into the global market, competing with African and European suppliers through Rome for WFP business. JLR has over 100 staff in Durban and others elsewhere in South Africa. JLR staff dismissed a project-related procurement of 233 MT of dry peas for Zimbabwe as too small to affect prices noticeably, so the same reaction can be expected to the 350 MT of


cowpeas it supplied to WFP in Mozambique. However, the world market for cowpeas is much thinner than that for dry peas: although annual global cowpea production for 2006 – 2008 was 3.3 million MT, the annually traded total for the same period averaged only 2,900 MT.88 The LRP volume then corresponds to 12 percent of world trade and may have had a small temporary impact on world-market prices.89 However, no time series data for cowpeas in Mozambique or South Africa was collected by the evaluation team.

In light of the assessment above, the evaluation team concludes that in all likelihood, the LRP project did not have an impact on the price of maize or cowpeas in procurement zones.

88 Cowpeas represented 7 percent of the total world production of 47 million tons legume crops in 2006 – 2008, i.e. 3.3 million tons. See: Akibode S and M Maredia (2011). Global and regional trends in production, trade and consumption of food legume crops, report submitted to SPIA, March 27, pp 10, 33 89 WFP’s project final report remains mute about the market impact of cowpea procurement.


12. NICARAGUA The Fabretto Children’s Foundation (FCF) implemented the LRP project in Nicaragua. FCF purchased a variety of commodities, including butter, cream, pinolillo, sugar, carrots, cheese, eggs, green peppers, onions, plantains, potatoes, sugar, tomatoes, and bananas.

The evaluation team’s analysis focuses on red beans, corn flour, cheese, green peppers, vegetable oil, and plantains, and is based on price data provided by FCF.

Accompanying the actual report, FCF provided a very comprehensive and detailed Excel file that the evaluation team used for analysis. The data are organized in groups of three observations: market price one week before, on the day of, and one week after procurement (Figures 12.1 and 12.2). The local market price for each time period indicates negligible variation. In addition, red bean and corn flour prices were extremely stable over the period. This coherence and degree of stability calls for caution in drawing conclusions about the market impacts of the corn flour and red bean purchases made under the project.

FIGURE 12.1: WHOLESALE BEAN PRICES MANAGUA

Source: Adapted from USDA Local and Regional Procurement Pilot Project, Evaluation Report –Final LRP Project Report Fabretto Children’s Foundation/Nicaragua Note: local market prices one week before procurement (blue line) and on the date of procurement (red line) are almost the same. The result is that the blue line is difficult to see.

800

900

1000

1100

1200

1300

1400

11/0

1/10

12/0

1/10

01/0

1/11

02/0

1/11

03/0

1/11

04/0

1/11

05/0

1/11

06/0

1/11

07/0

1/11

08/0

1/11

09/0

1/11

USD

/MT

Wholesale Bean Prices, Managua (USD/MT)

Local Market Price (oneweek beforeprocurement)

Local Market Price(date of procurement)

Local Market Price (oneweek afterprocurement)


FIGURE 12.2: WHOLESALE CORN FLOUR PRICES, MANAGUA

Source: Adapted from USDA Local and Regional Procurement Pilot Project, Evaluation Report –Final LRP Project Report Fabretto Children’s Foundation/Nicaragua Note: local market prices one week before procurement (blue line) and on the date of procurement (red line) are almost the same. The result is that the blue line is difficult to see.

The evaluation team examined data for the other commodities purchased by the project, but the prices of plantains and vegetable oil contained significant errors. As such, the team decided not to use these data.

Prices before, during, and after procurement for cheese and green peppers were also fairly coherent over time, but variations within the two week period do not show a clear pattern. At times, prices vary before the purchases, but they also vary after them in other instances. Without analyzing more data, it is impossible to conclude whether the procurements have had any impact on prices.

As such, the evaluation team finds it unlikely, that LRP had any impact on market prices for the purchased commodities in Nicaragua.

USD

/MT

Wholesale Corn Flour Price, Managua

Local Market Price(one week beforeprocurement)

Local Market Price(date of procurement)

Local Market Price(one week afterprocurement)


13. NIGER CRS and Mercy Corps separately implemented two LRP projects in Niger. Each is addressed in turn.

13.A. Niger CRS CRS implemented a local procurement project in parts of Tillabéri and Zinder Regions in Niger from March to September 2011. The project provided vouchers that allowed beneficiaries to obtain 4,760.5 MT of cereals, 793 MT of cowpeas, and 58.4 MT of vegetable oil over a four-month period.

The evaluation team visited the project in August 2011 and based its assessment on data collected with various project stakeholders during this visit.

A total of 38 local vendors exchanged vouchers denominated in CFA francs for staple foodstuffs including millet, sorghum, beans, vegetable oil, rice, manioc flour, and maize in one market in Tillabéri Region and in five markets in Zinder Region. CRS generally distributed vouchers the day before the local weekly market at biweekly or monthly intervals from May 2011 to August 2011.

CRS followed market price trends in voucher-redemption markets as well as control markets nearby. In Tillabéri Region, CRS monitored one redemption market (Ouallam) and two control markets; in Zinder Region, it monitored five redemption markets and two control markets.

In contrast, CRS did not track price trends along the market chain that led to the marketplaces where the beneficiaries redeemed the LRP vouchers because “producers in Niger sell commodities in the months immediately after the harvest (October) and not during the period of project intervention.” In other words, by the start of the next calendar year, farmers in Niger have sold almost all their marketable surpluses of these commodities and those who will supply them to markets in the May to August period will be traders. Therefore the LRP project-induced higher prices could not have elicited greater supply from Nigerien90 farmers.

Two additional observations to the analysis in the last paragraph can be made. First, Nigerien farmers were not the only producers of these commodities: Nigerian farmers produced the majority of the millet, sorghum, maize and manioc flour sold to LRP beneficiaries in Zinder Region; farmers from Burkina Faso, Togo, and/or Benin produced the majority of these same commodities distributed to LRP beneficiaries in Tillabéri Department (part of Tillabéri Region); and the rice and vegetable oil procured came mostly from the regional market. Among the foodstuffs listed, Niger is a net exporter of only beans (cowpeas) and during the May to August period, Niger may have had to reimport even these.

Second, any market impact due to the procurement does not directly affect Nigerien farmers at that time, but this point might omit other possible market impacts, which the additional purchasing power spread over the four month voucher-redemption period may have. Traders (mostly outside Niger), who bought these commodities in the immediate post-harvest period and stored them for several months, may make greater profits (or conceivably lesser losses) because of the higher demand due to the LRP project. However, the further this trade is located from the Nigerien redemption markets, the smaller this impact would be. In addition, some of this additional profit would have been passed to other traders along the market chain between these traders and the targeted LRP project beneficiaries. Evidence from Zinder Region indicated that some additional profit accrued to the Nigerien vendors who participated in voucher redemption: they were enthusiastic about the scheme, stated that it had significantly raised their throughput, and told the evaluation team that they had recruited additional staff, albeit temporarily, to deal with the extra market transactions.

Third, CRS did not identify the sources of the foodstuffs, either within or outside Niger for which higher prices due to a brief period of additional (LRP) purchases might be expected. Though they were able to 90 Note that “Nigerien” means “from Niger”; “Nigerian” means “from Nigeria”.


measure market impact at markets within Niger, they were not able to measure the impact close to the non-Nigerien purchase markets. To do so, CRS would first have had to identify the procurement zones outside Niger, then follow-up quickly to get data from local markets. Where these data were not available, CRS would have had to collect primary data specifically for its own assessment needs. If non-Nigerien procurement zones are multiple for several commodities, the data collection exercise comes to require significant cost and effort. The total cost of assessing price impact of procurement, within Niger and outside it, might be as much as an order of magnitude greater than it was for distribution. If catchment zones of the production and redemption markets overlap in space and time, the data collection becomes simpler because it would all take place within Niger. However, interpreting the analysis would be potentially more complicated as the first effect tends to raise prices while the second tends to suppress them, with the possibility of no price change masking equal and opposite non-zero market effects.

Fourth, despite the CRS argument that Nigerien farmers would themselves have directly sold little or no commodities to the beneficiaries through the project’s voucher mechanism that was in operation 7 – 11 months after the previous harvest, there is the possibility that farmers would alter their decisions for the following agricultural season, about what and how much to plant and/or about investment in agricultural inputs in light of the prices observed for the LRP project commodities. Farmers around the redemption markets might see lower prices for these commodities due to an LRP project-induced increase in supply and lower prices at the start of the planting time (June) for cereals and cowpeas. This would lead those who habitually produce surpluses to sow lower acreages of one of the LRP project commodities for the previous agricultural year or switch to non-LRP crops with prices unaffected by the LRP project. Conversely, farmers around the collection markets feeding the market chain that leads to LRP project redemption markets who enjoy some flexibility in the allocation of their acreages and inputs may act to increase production of LRP project commodities, prices for which they saw rise at planting time.

Fifth, CRS’ USDA-financed LRP project purchases were a very small fraction of the throughput in well-integrated agricultural markets in the eastern Sahel; therefore the working hypothesis must be to assume no impact unless proven otherwise.

In addition, CRS staff ensured that approved traders who redeemed vouchers had enough food to meet demand, so it was not a question of increased demand meeting a constant supply.

In summary, the most important conclusion is that in the absence of price data from known or suspected producing zones for the foodstuffs for which LRP project beneficiaries exchanged their vouchers, it is not possible to demonstrate convincingly the presence or absence of market impact of the procurement process for CRS’ LRP project.

In the case of voucher projects, identifying the production zones or even actual farms would be nearly impossible, when the producers are hundreds of thousands of West African smallholder farmers.

Had CRS identified the production zones, it would have been too expensive to monitor prices for a group of commodities converging on several redemption markets—each paired with a control market— across multiple market chains (many of which cross national boundaries).

To pursue this evaluation in a practical fashion, it would be appropriate to make several assumptions to severely prune down the number of market chains considered, and perhaps only for the more important of the commodities in question. Moreover, to the extent possible, the data collection should take place only at key points along relevant market chains. For instance, if the procurement takes place during a season when producers have long since sold their marketable surpluses, then – for that year – any market impact lies further down the market chain. However, in the longer term, if producers and/or traders see higher prices in the period after the producers’ normal sales period, they may act to increase production during the next agricultural year, with a consequent impact on next season’s prices. Particularly in this case, the diffuse impact over subsequent years would probably be so small that it would be undetectable.


A rise in the price of a food commodity helps the producer and penalizes the consumer. Whether it is a favorable outcome is subjective and depends on the weightings assigned in analysis to the producer and the consumer. In a producing zone, higher prices may (a) help smallholders raise incomes and thus increase their ability to invest in future increased agricultural production and (b) provide a motivation to invest in this increased production. In the long run, the boosted agricultural production may play an important role in rural economic development. However, even in major food-producing areas there exist food-deficit households. Lower prices will reduce food bills in households that consume more than they produce, which is of particular short-term importance during famine.

The above discussion leads the evaluation team to conclude that no evidence points to an impact of the LRP pilot in Niger on the prices of the procured commodities.


13.B. Niger Mercy Corps In FY 2011, Mercy Corps implemented a separate USDA-financed LRP project in Niger. It took place in parts of the Filingué Department of Tillabéri Region, with food-aid distributions that were adjacent to CRS’ distributions in the Ouallam Department of the same region. Mercy Corps procured and distributed 4,699 MT of maize and 472 MT of cowpeas, as well as millet, cooking oil and salt,91 to 78,608 beneficiaries. Each beneficiary household received rations calculated to meet 80 percent of its monthly needs for five months during the tail end of the 2009 – 2010 hungry season (September 2010) and for four months during the 2010 – 2011 hungry season (April – July 2011). Competitive procurement of maize, cowpeas and millet for these distributions took place in September 2010, March 2011 and July 2011, with about three-quarters of the total procured in March 2011.

The evaluation team’s analysis of the potential market impacts connected to this project rests on work done by Cornell University as well as data collected by the team from key stakeholders in the field.

Mercy Corps benefited from market impact analysis by Cornell University. Cornell performed multiple regressions to establish price variations attributable to local purchases of maize, millet92 and cowpeas in a moderate-sized market in the town of Filingué (near the distribution sites) and in other Nigerien markets. The regression analysis controls for variables such as seasonality, rainfall patterns, inflation, transport costs and – importantly during a period characterized by sharply rising world food prices – global market prices. The authors sought statistically discernible price differences attributable to the procurements and distributions. Though the traders purchased the maize (and to some extent the millet and cowpeas) from outside Niger in order to meet the terms of the Mercy Corps contracts, their analysis covers only procurement effects within Niger.

The analysis finds that the procurement of millet causes statistically insignificant changes in the millet price; the authors attribute this small change not only to LRP financed by USDA but also due to WFP procurement taking place at the same time. Similarly, they note that procurement of cowpeas shows no statistically significant effect. However, the authors find that Mercy Corps’ local purchases of maize appear to have immediately increased maize prices by 6.7 percent, by 11.6 percent after one month, falling to 8.8 percent after two months. This appears plausible as the maize contracts accounted for about 80 percent of the commodities purchased under the project. While Mercy Corps procured the maize through Nigerien traders, most, if not all, of these traders purchased the maize in Benin and Mercy Corps did no market research in Benin to establish market conditions. In particular, it is not known if other donors were purchasing maize there as food aid for Niger or for other countries at a time of significant residual food insecurity following the drought of 2009.

To the extent that procurements cause significant but short-lived price rises, followed by prices dropping again by roughly the same amount, they contribute to increased price volatility, though there is no reason to believe that the fluctuations continue after the initial price spike. Cornell University found a small statistically significant increase in the volatility of cowpea and maize prices attributable to Mercy Corps’ distributions but none attributable to procurement (either on the part of Mercy Corps alone for maize or cowpeas, or due to procurements made by Mercy Corps and WFP, for millet).

In conclusion, the evaluation does not consider an impact on prices caused by the LRP project to be probable.

Mercy Corps’ calculations of the Import Parity Price (IPP) show that IPP equals the local market price for all commodities, other than salt, over the procurement period. 91 Mercy Corps allocated the oil and salt through a voucher scheme. However, as the voucher was for a fixed volume of oil and a fixed weight of salt and as a single approved local trader honored the voucher, the entire food aid mechanism was essentially a distribution of fixed rations per household. 92 Mercy Corps used the millet to restock local cereal banks, which did not immediately sell any of it because the distributions at the household level of maize and cowpeas largely met immediate needs. Therefore Barrett et al. (2011) assumed that the millet distributed would have no market impact during the period of analysis. They also decided not to examine the impact of the distributions of cooking oil or salt.


14. PAKISTAN WFP implemented an LRP project in Pakistan as an emergency feeding project in response to the monsoon-related floods that destroyed much of the cropland and crops in parts of Pakistan in 2010. For this project, WFP procured 10,062 MT of wheat locally from the Government of Pakistan through the Pakistan Agricultural Storage & Services Corporation Ltd. (PASSCO).

WFP provided no time series or market throughput data for this LRP project to enable the evaluation team to perform an independent analysis of market impacts due to the procurement of this wheat. However, WFP has provided their brief assessment of this issue in their final project report, as follows:

“Sizeable wheat stock balances have remained available in Pakistan, despite the recent recurrence of floods in Sindh and Balochistan, ensuring that the wheat balance sheet continues to be in surplus. As such, and in the context of prevailing economic hardship, the local purchase of wheat by WFP did not only have no real impact on prevailing market prices – due to the comparatively small scale of transactions – but also helped to ease the domestic market and provide some support to the domestic economy.”

In the absence of any of the data needed to perform an independent analysis of market impact from the LRP project, the evaluation team cannot exclude the possibility that such an impact occurred. However, the explanation offered in WFP’s report makes impact appear unlikely.


15. TANZANIA WFP implemented an LRP project in Tanzania as part of its P4P initiative. This local procurement project procured 2,475 MT of maize and 424 MT of beans from Savings and Credit Cooperatives (SACCOs) within Tanzania for distribution to other areas within that country. Table 15.1 provides a summary of the procurement dates from the report submitted for this project.

The evaluation team assessed the impact of WFP procurements on the maize market through analysis of price trends in markets where procurement took place; for a similar assessment of beans, data were not available.

WFP provides maize price information for eight main SACCO procurement markets: Kware, Didihama, Usomama and Gallapo (located near Arusha) and Soko Kuu, Mrijo, Kwa Motoro and Kibaigwa (located near Dodoma), some of which appear in Table 15.1, which shades the Arusha markets in yellow and Dodoma markets in orange. WFP collected the price information twice monthly from May to November 2010, with some gaps in the data.

TABLE 15.1: PROCUREMENT REPORTS FOR THE WFP TANZANIA LRP PROJECT

Quantity (MT)

Contract signature date

Delivery dates Market zone

SACCO market

247 1/17/2010 NA Arusha Kware

30 3/19/2010 NA Arusha Usomama

50 08/06/2010 12/16/2010 Dodoma Didihama

100 08/06/2010 NA Manyara NA

227 08/06/2010 10/15/2010 – 11/9/2010 Dodoma Soko Kuu*

100 08/11/2010 10/13/2010 Dodoma Mrijo*

468 10/05/2010 NA Dodoma Kibaigwa

200 10/14/2010 NA Singida NA

100 12/18/2010 NA Arusha NA

100 01/01/2011 NA Dodoma NA

50 1/14/2011 NA Moshi NA

802 02/02/2011 NA Rukwa NA

2475 TOTAL Source: Adapted from USDA Local and Regional Procurement Pilot Project, Evaluation Report –Final LRP Project Report WFP/Tanzania Notes: 1. these two asterisked (‘*’) markets may be inverted. 2. NA = not available

WFP also forecast prices for Arusha and Dodoma, based on a 15-year seasonal index and other factors. Figure 15.1 superimposes the forecast for Arusha and the price levels actually recorded in the four northern markets located near Arusha. Price data are not available for some months. Figure 15.2 presents the same information for Dodoma (forecast) and nearby markets (actual prices).


FIGURE 15.1 TANZANIAN MAIZE PRICES AND ARUSHA FORECAST PRICE, 2010

Source: Adapted from USDA Local and Regional Procurement Pilot Project, Evaluation Report –Final LRP Project Report WFP/Tanzania Note: breaks in the lines reflect gaps in the data, for which price data were not available.

FIGURE 15.2: TANZANIAN MAIZE PRICES AND DODOMA FORECAST PRICE, 2010

Source: Adapted from USDA Local and Regional Procurement Pilot Project, Evaluation Report –Final LRP Project Report WFP/Tanzania Note: breaks in the lines reflect gaps in the data, for which price data were not available.

150

200

250

300

350

Tanz

ania

n sh

illin

gs p

er k

ilogr

am

Kware

Didihama

Usomama

forecast

150

200

250

300

350

Tanz

ania

n sh

illin

gs p

er k

ilogr

am

Soko Kuu

Mrijo

Kibaigwa

forecast


In Figures 15.1 and 15.2, forecast prices were similar and correctly anticipated the general price level. However, neither correctly matches the month-to-month price variations in the individual markets, with the exception of Mrijo and Soko Kuu in October-November: neither is generally accurate. However, they do serve to remind us of the average seasonal profile of maize prices.

Table 15.1 provides details of maize procurements from Kware, Didihama and Usomama, price trends for all of which appear in Figure 15.1. However, the contract signature dates for Kware and Usomama occurred before the start of the maize price data collection in these markets. Given the two to three month lag between contract signature and delivery, it is possible that some market impact occurred in these markets in May, though the larger procurement (247 MT from Kware) would probably have been delivered by then and the smaller procurement (30 MT from Usomama) would be expected to have a smaller effect on prices. In any case, price data for April to compare to May prices are lacking. However, from May to June, prices rose in Kware (contrary to seasonal norms as shown in the forecast) and fell in Usomama. For Didihama, Table 15.1 shows a June signature date for a contract for 50 MT of maize and a December delivery date. Though the price was flat from June to July after signature of the contract, it rose from July to August, contrary to the normal seasonal trend. However, the time profile of purchases that culminated in the eventual delivery in December is not known, so it is difficult to know where to look for a price spike, particularly in the absence of data for September and December.

Figure 15.2 is also drawn from Table 15.1 and shows 2010 maize price trends for Soko Kuu, Mrijo and Kibaigwa. August contracts were concluded for Soko Kuu (227 MT) and Mrijo (100 MT) but no price data are available for that month or for September. Delivery from Mrijo took place in October and from Soko Kuu in October-November at which point their prices were identical and coincided almost exactly with the forecast trend (which had not taken LRP into account). The contract signature date in Kibaigwa was October, but this did not provoke higher prices in November.

In conclusion, although the quantity of data for maize market impact evaluation is low, the available evidence makes market impact appear unlikely.

No corresponding data in either the WFP report or the associated data file allow evaluation of market impact for the procured beans. Neither were there market throughput data with which to perform a limited elasticity analysis. The report contains various data on the five procurements of beans, all from the Kagera region, with deliveries between March 31, 2011, and September 10, 2011. The only prices noted are the price paid per MT for each of the procurements, and there are too few data points in this figure to draw firm conclusions. Moreover, the prices are those paid by WFP, not the prices of beans in the local markets; market price data for beans were not available to the evaluation team for this analysis. For two of the procurements, both contracts were signed on October 11, 2010, and both deliveries were made to WFP warehouses on March 31, 2011, but the prices paid by WFP per MT of the beans were different (690 TZS/MT and 765 TZS/MT). This small assessment is further confounded by the fact that the two procurements that were delivered on March 31 were direct purchases whereas the other three procurements were by competitive tenders. Thus, conclusions about the market impact of WFP’s bean procurement in Tanzania cannot be drawn.


16. UGANDA World Vision implemented an LRP project in Uganda, where it procured maize meal, cassava flour, millet flour, beans, rice, sorghum, fortified and unfortified vegetable oil, and maize grain, all purchased from a total of 162 approved vendors in Uganda. World Vision distributed all commodities in Uganda through the voucher program. Some beneficiaries received vouchers as part of a food for work scheme, while others who were unable to work received vouchers without conditions. Beneficiaries redeemed the vouchers in identified markets in exchange for small quantities of food over months and not in bulk amounts on a few days.

The evaluation team’s assessment is based on analysis done by Cornell University for World Vision and analysis of World Vision data.

In a narrow sense, the voucher approach conflates project-related procurement with project-related distribution, as the beneficiary redeems the voucher directly in the local market where approved traders offer an approved range of foodstuffs in return. However, the “procurement market” in a wider sense probably lies elsewhere if, as is often the case, the approved traders source the commodities elsewhere in the country or the region.

World Vision prepared a project report incorporating market impact analysis done by a Cornell University team in support of the Learning Alliance. The Cornell team used econometric methods for their market impact analysis and found no significant effect on prices of maize or sorghum following voucher distribution. However, the report notes:

“World Vision voucher distributions were associated with statistically significant, temporary decreases in local market prices of maize flour (-13.9 and -6.3 percent in the month of and following voucher distribution, respectively) and dried beans (-16.0 percent in the month of voucher distribution, before falling to a statistically insignificant level the following month).”

At the same time, Cornell’s researchers suspect that those changes were likely caused by exogenous but unobservable factors, given the very small proportion of voucher redemptions (less than 5 percent, according to the Cornell report).

The Cornell report goes on to suggest that concerns about changes in price brought about by relatively small volumes of maize flour and dried beans are misdirected. WFP procurements were much more substantial, and showed no market effect. The report states: “WFP’s quite substantial procurements of maize grain in Uganda (which averaged nearly 11,000 MT during months of procurement) were not statistically significantly associated with maize grain market prices.”

World Vision created a database of market price data for procured commodities, collected from a number of vendors in multiple markets, towns, and provinces over a large number of dates, covering part of 2010 but primarily 2011. The database contains prices by market level (wholesale or retail) and gives, among other details, the date of collection. However, coverage is not continuous.

The project distributed food vouchers in eight tranches from December 2010 to August 2011, as shown in Table 16.1. For each round, the evaluation team assumed that voucher purchases by the beneficiaries took place 10 days after the midpoint of the voucher-distribution period.93 Beneficiaries were free to choose the market in which they redeemed their vouchers and the project report gives no additional data to identify during which portion of a given distribution period beneficiaries were likely to redeem their vouchers in a given market. Further, as beneficiaries were free to choose within a basket of permitted commodities, the estimated distribution date applies to all such commodities. Finally, for lack of data on the purchase of the commodities by the vendors, the evaluation team used distribution dates as proxy for procurement dates to identify price impacts in procurement markets.

93 The World Vision Acholi voucher project evaluation report notes: “Most beneficiaries redeemed their vouchers with 10 days” (page 4).


TABLE 16.1 DATES OF ACHOLI VOUCHER PROJECT’S FOOD-VOUCHER DISTRIBUTION TRANCHES AND DISTRIBUTION DURATIONS Round 1 Round 2 Round 3 Round 4 Round 5 Round 6 Round 7 Round 8 Dec 10-22, 2010

Jan 17-26, 2011

Feb 8-15, 2011

Mar 23-Apr 6, 2011

Apr 27- May 4, 2011

June 6- 15, 2011

June 28-July 7, 2011

July 30-Aug 11, 2011

13 days 10 days 8 days 14 days 8 days 10 days 10 days 8 days Source: Adapted from USDA Local and Regional Procurement Pilot Project, Evaluation Report –Final LRP Project Report World Vision/Uganda Acholi Voucher Program The evaluation team limited the analysis to three important eligible commodities: beans, sorghum and maize meal. Examination of the data revealed that for each commodity the greatest number of price data occurred for agwede beans, sorghum and “maize flour 1” at the retail level for both Pader/Agogo market (distribution zone) and Lira (procurement market). Consequently, these three commodities for these two markets were retained for analysis.94

First, the evaluation team generated monthly average prices, shown in Figures 16.1 – 16.6. These figures give a broad overview of the price trends over the period under consideration. A. Pader/Agogo distribution market

FIGURE 16.1: RETAIL BEAN PRICES IN PADER, UGANDA

Source: Adapted from USDA Local and Regional Procurement Pilot Project, Evaluation Report –Final LRP Project Report World Vision/Uganda Acholi Voucher Program Note: Breaks in the lines reflect gaps in the data, for which price data were not available.

94 Although the team considers procurement and distribution to coincide, Pader was one of the markets where beneficiaries were likely to redeem their vouchers, while Lira was one of the markets in an area of less food stress and more likely to be a source market.


FIGURE 16.2: RETAIL MAIZE-FLOUR PRICES IN PADER, UGANDA

Source: Adapted from USDA Local and Regional Procurement Pilot Project, Evaluation Report –Final LRP Project Report World Vision/Uganda Acholi Voucher Program

FIGURE 16.3: RETAIL SORGHUM PRICE IN PADER, UGANDA



B. Lira procurement market

FIGURE 16.4: RETAIL BEAN PRICES IN LIRA, UGANDA


FIGURE 16.5: RETAIL SORGHUM PRICES IN LIRA, UGANDA



FIGURE 16.6: RETAIL MAIZE-FLOUR PRICES IN LIRA, UGANDA


Table 16.2 summarizes the change in price corresponding to the distribution (Pader) or procurement (Lira), in terms of the direction of the price change to the month in question from the previous month, as shown in Figures 16.1 – 16.6. For each distribution tranche, Table 16.2 shows (1) the corresponding distribution period and (2) the corresponding estimated date of procurement or distribution (following the methodology explained above). Following this information, for each market-commodity pair, the monthly average price for the estimated distribution price is compared to the monthly average price for the previous month, where data exist, to establish the direction of change of price.95 There are many missing cases, corresponding to the months without data in Figures 16.1 – 16.6. Inspection of the figures reveals that, where data allowed determination of the direction of change, commodity prices rise during the months leading up to the July-August harvest period (shaded in green), as would be expected. Moreover, there are no notable price spikes in months following procurements/ distributions. This analysis therefore reveals no suggestion of market impact due to LRP project activities.

95 This approach duplicates the February results in table 2 because the team’s methodology estimates that two distributions/procurements (corresponding to tranches 2 & 3) would have taken place in February. Table 3 replicates this duplication.


TABLE 16.2: MONTHLY PRICE CHANGES CORRESPONDING TO EACH “ACHOLI VOUCHER PROJECT” DISTRIBUTION/PROCUREMENT (COMPARING MONTHLY

AVERAGE PRICES)

Voucher tranche 1 2 3 4 5 6 7 8

Distribution period

Dec 10-22

Jan 17-26

Feb 8 -15

Mar 23-Apr 6

Apr 27- May 4

June 6- 15

June 28-July 9

July 30-Aug 11

Estimated date of procurement/distribution

27th Dec

1st Feb

22nd Feb 9th Apr

11th May

21st Jun 13th Jul

15th Aug

Pader (distribution)

Beans nd P↑ P↑ P↑ P↑ P↓ P↓ nd Maize Flour nd P↑ P↑ P↑ nd nd nd nd Sorghum nd P↑ P↑ P↑ P↑ P↑ P↑ nd

Lira (procurement)

Beans nd nd nd nd P↑ P↓ P↓ nd Maize Flour nd nd nd nd P↑ P↑ P↓ nd Sorghum nd nd nd nd P↑ P↑ P↑ nd

Source: adapted from USDA Local and Regional Procurement Pilot Project, Evaluation Report –Final LRP Project Report World Vision/Uganda Acholi Voucher Program Notes: grey shading = harvest period; “nd” = no data, i.e. data permitting calculation of direction of monthly price change were lacking; beans are agwede beans; maize flour is “maize flour 1”; P↑ and P↓ show the direction of price changes relative to the previous month. The evaluation team repeated the analysis using the same data in their unprocessed form, capturing changes between individual dates96 on either side of the estimated date of procurement/distribution, rather than differences in monthly averages, using the closest specific dates on either side of the estimated distribution/procurement date (as long as individual pairs of dates compared fell in successive months, rather than further apart in time). This more fine-grained analysis uses (1) data that are, on average, closer to the estimated procurement/distribution date but (2) fewer data overall. Like the previous analysis, this analysis also suffers from missing data. Table 16.3 provides the results.

96 Sometimes there were several price data for a given date, in which case the mean of these was used.


TABLE 16.3: PRICE CHANGES CORRESPONDING TO EACH “ACHOLI VOUCHER PROJECT” DISTRIBUTION/PROCUREMENT

(USING PRICE DATA FOR SPECIFIC DATES)

Voucher tranche 1 2 3 4 5 6 7 8

Distribution period

Dec 10-22

Jan 17-26

Feb 8 -15

Mar 23-Apr 6

Apr 27- May 4

June 6- 15

June 28-July 9

July 30-Aug 11

Estimated date of procurement/distribution

27th Dec

1st Feb

22nd Feb 9th Apr

11th May

21st Jun 13th Jul

15th Aug

Pader (distribution)

Beans nd P↑ P↓ P↓ P↑ P↓ P↓ nd Maize Flour nd P↑ P↓ P↑ nd nd nd nd

Sorghum nd P↑ P0 P0 P↑ P↑ P↑ nd

Lira (procurement)

Beans nd nd nd P↑ P↑ P↓ P↓ nd Maize Flour nd P↓ nd nd P↑ P↑ P↓ nd

Sorghum nd nd nd nd P↑ P↑ P↑ nd Source: adapted from USDA Local and Regional Procurement Pilot Project, Evaluation Report –Final LRP Project Report World Vision/Uganda Acholi Voucher Program

Notes: grey shading = harvest period; “nd” = no data, i.e. data permitting calculation of direction of monthly price change were lacking; ; P↑ and P↓ show the direction of price changes relative to the previous month; “P0” = no price change; beans are agwede beans; maize flour is “maize flour 1.” The results of Table 16.3 broadly echo those of Table 16.2, with six exceptions. Table 16.3 shows that prices remain the same before and after distributions 3 and 4 for sorghum in Pader market whereas Table 16.2 shows price rises in each case; in Table 16.3 the price of maize flour falls in Lira following tranche 2 and that of beans rises following tranche 4, whereas there are no data to allow a determination in Table 16.2; and the price of beans in Pader market following tranches 3 and 4 falls in Table 16.3, whereas they rise in Table 16.2. In the absence of voucher purchases, we would expect prices to rise towards harvest, which takes place in approximately July and August. They do so from January to May in Table 16.2, after which prices fall in the bean market, perhaps in anticipation of a good bean harvest. If the additional demand generated by vouchers in the markets had any effect it would be to raise prices further, particularly in the markets in which beneficiaries redeem them. However, Table 16.3 notes price falls associated with purchases, and they are disproportionately in Pader, a distribution market, at the centre of any impact of the additional demand. This is difficult to explain and suggests that these price changes are spurious results of limited and sporadic data or perhaps other market factors. The two methodologies used do not yield identical results but neither provides evidence of price spikes attributable to LRP or price rises that lie outside what we may consider normal seasonal price profiles for the commodities and markets considered. The evidence for a market impact caused by the LRP project is lacking. In conclusion, considering the results of the Cornell University report and the evaluation team’s analysis of the World Vision data, there is no evidence to negate the null hypothesis of no price impacts due to the LRP project.


17. ZAMBIA Land O’ Lakes implemented a local procurement project in Zambia using various commodities procured and distributed in Zambia. All procurements were made in Lusaka between December 2010 and September 2011, comprising 2,745 MT of maize meal, 124.5 MT of maize grain, 500.48 MT of beans, 173.68 MT of vegetable oil, and 138.05 MT of High Energy Protein Supplement (HEPS).

The evaluation team used analysis based on supply elasticities, as well as analysis performed by Cornell University.

Cornell University analyzed market prices associated with Land O’ Lakes’ procurements and found significant impacts for some commodities and little or no price impact for others. The analysis can be found in Land O’Lakes’ final project report, and states:

Land O’ Lakes’ local purchases…of beans under the USDA LRP pilot program had no economically or statistically significant effect on retail bean prices in Zambia. By contrast, Land O’ Lakes’ procurement of maize meal had statistically significant effects on retail food prices…. [W]e estimate that maize meal procurement caused breakfast meal prices to fall by 6.4 and 6.6 percent the same month as procurement and distribution, respectively. Those price effects further decrease to -11.2 to -12.5 percent in the month following procurement before vanishing two months later to small point estimates not statistically significantly different from zero. Similarly, roller meal prices fell an estimated 9.9 percent in procurement month, further decreasing to -16.1 percent the month following procurement, before settling back to no economically or statistically significant effect two months after procurement. Given that these regressions already control for usual seasonal patterns on prices, the apparent negative retail price effects of maize meal procurement by Land O’ Lakes likely reflect a combination of expended production of maize meal by prospective suppliers above and beyond Land O’ Lakes’ purchases, as well as reduced consumer demand by recipients who received maize meal rations, with the increased market supply and decreased market demand jointly, albeit only temporarily, driving down prices.

Cornell’s regression analyses estimated smaller effects for procurement market price impacts in procurement markets. Prices for vegetable oil were also found to have decreased by 21.5 percent in the month of distribution, though these effects were not found to be statistically significant in distribution markets. It should be noted that purchase quantities for all commodities procured under the LRP project, however, were less than 1 percent of total production, which indicates that caution must be used in interpreting the possible market impact of the LRP project purchases on their own.

Importantly, the Zambian Food Reserve Agency (FRA) was highly active in making maize purchases and sales during the time period of the project, which also coincided with a highly contested presidential election. Data from FRA were unavailable to both Cornell and the evaluation team, precluding estimation of the effect of their price floor and aggressive purchases from farmers.

The Cornell University report thus concludes that, for some commodities, regressions revealed price impacts but, given that they are negative, these impacts are, in fact, unlikely to be due to the procurements made for the LRP project. To the extent that the procurement and distribution zone markets are tightly linked, the reduced demand from LRP beneficiaries may have been communicated to the procurement zone markets and dampened prices there, though it seems unlikely that the net result would be a statistically significant negative effect. A more likely (albeit unquantified) factor seems likely to be the pre-election activity of the FRA. In sum, given the level of uncertainty surrounding this analysis, it is possible that the LRP project may have had negligible impacts or none at all.

In lieu of an econometric analysis, the evaluation team compares the available price series to the readily available evidence. Firstly, the Zambian agricultural calendar, shown in Figure 17.2, indicates that the main harvest begins in early April from which point crop prices tend to drop as harvested crops are sold. Prices


normally begin to rise again in July and reach their peak in November to February (the “lean” or “hungry” season in Zambia).

FIGURE 17.2: ZAMBIAN AGRICULTURAL SEASON CALENDAR

Source: USDA Local and Regional Procurement Pilot Project, Evaluation Report –Final LRP Project Report Land O’Lakes/Zambia.

Figure 17.3 shows wholesale and retail prices of mealie-meal (maize meal) in Lusaka Central Market over the procurement period, with the linear trend for each. The downward trend in prices across this period is consistent with the pattern that the Zambian agricultural calendar would lead one to expect.

FIGURE 17.3: PRICE TRENDS FOR MAIZE MEAL IN THE LUSAKA CENTRAL MARKET


Figure 17.4 provides wholesale and retail prices for beans in the same market and over the same period. The prices follow a similar downward trend as that of maize, in tune with the agricultural calendar. Thus for maize

Price for Mealie-Meal (Maize Meal) in Lusaka Central Market

900950

1,0001,0501,1001,1501,2001,2501,3001,3501,400

14-D

ec-10

23-D

ec-10

25-Ja

n-11

22-Feb

-11

7-Mar-1

1

22-M

ar-11

5-Apr-1

1

19-A

pr-11

5-May-1

1

23-M

ay-11

8-Jun-11

28-Ju

n-11

13-Ju

l-11

26-Ju

l-11

8-Aug-11

22-A

ug-11

5-Sep-1

1

19-S

ep-11

Date

Pric

e (Z

MK

/kg)

Retail Wholesale Linear (Retail) Linear (Wholesale)


and beans the general price trend is consistent with the causal influence expected from the underlying seasonality, though it cannot be concluded that this is the only, or even the main, causal factor.

FIGURE 17.4: PRICE TRENDS FOR BEANS IN THE LUSAKA CENTRAL MARKET


Land O’ Lakes purchased 2,745 MT of maize and maize meal over nine months. Zambia’s annual maize production is 2.8 million MT, of which about a quarter, 700,000 MT, is marketed. Maize is assumed to be consumed within a year of harvesting and to change hands three times over that period, leading to 2.1 million MT maize transactions annually, or market throughput of 175,000 MT monthly. The short-run price elasticity of supply for maize (0.23) gives an estimate of the impact of the initial extra purchase on the market price of +2.3 percent. Inspection of Figure 17.3 shows that, given the scale of price fluctuations of the order of 5 percent around the linear trend, such a small LRP-induced change would not be noticed. As the procurements resulted in fairly regular subsequent deliveries of the same magnitude, market expectations would begin to adjust to these and the elasticity of supply would rise over time. However, within a nine-month period, full long-run adjustment would not be expected. Finally, the cessation of LRP purchases in September 2011 would have caused a short-term price change of -0.05 percent. Available price data do not extend beyond early September 2011 but, extrapolating from the price profile in Figure 17.3, it can be inferred that it is unlikely that this price change would have been noticeable either.

Similar logic applied to Zambia’s bean market yields a projected short-term price rise due to a procurement of 21 percent, based on an annual production of 30,000 MT of beans, of which 42 percent (12,600 MT) are marketed and a short-run price elasticity of supply of +0.25. This is significantly greater than the price rise projected for maize. This would stand out among the fluctuations of the order of 5 percent apparent around the Lusaka Central Market bean price trend line in Figure 17.4, so this additional impact could be detected.

Figure 16.6, for cooking oil, shows a slightly increasing price trend over the same period. However, Zambia imports around 60,000 MT of cooking oil annually,97 and this commodity’s price should not be expected to 97 “Zambeef boosts palm oil plantation”, 5th January 2011, http://ir.zambeefplc.com/profiles/investor/ResLibraryView.asp?BzID=1988&ResLibraryID=42901&Category=1788

Price for Beans in Lusaka Central Market

4,000

5,000

6,000

7,000

8,000

9,000

10,000

11,000

12,000

14-Dec-10

23-Dec-10

25-Jan-11

22-Feb-11

7-Mar-11

22-Mar-11

5-Apr-11

19-Apr-11

5-May-

11

23-May-

11

8-Jun-11

28-Jun-11

13-Jul-11

26-Jul-11

8-Aug-

11

22-Aug-

11

5-Sep-

11

Date

Pric

e (Z

MK

/kg)



follow the dictates of the agricultural calendar. LRP project-related purchases over nine months equaled 175 MT, or 20 MT monthly. This corresponds to a 0.4 percent rise in purchases. A short-term supply elasticity of 0.40 leads to a predicted price rise of 0.01 percent, unidentifiable among the other fluctuations in cooking-oil prices in Figure 17.5.

FIGURE 17.5: PRICE TRENDS FOR COOKING OIL IN THE LUSAKA CENTRAL MARKET


Prices for Cooking Oil in Lusaka Central Market

8,000

9,000

10,000

11,000

12,000

13,000

14,000

15,000

16,000

14-Dec-10

23-Dec-10

25-Jan-11

22-Feb-11

7-Mar-11

22-Mar-11

5-Apr-11

19-Apr-11

5-May-

11

23-May-

11

8-Jun-11

28-Jun-11

13-Jul-11

26-Jul-11

8-Aug-11

22-Aug-11

5-Sep-11

Date

Pric

e (Z

MK

/kg)



FIGURE 17.6: RETAIL PRICE TREND FOR HIGH ENERGY PROTEIN SUPPLEMENT (HEPS) IN LUSAKA CENTRAL MARKET


Figure 17.6 shows the decreasing price trend over the period of interest for HEPS. Although this commodity follows a decreasing price trend, like maize and beans, this is not due to Zambian climatic seasonality. Price elasticities of supply for HEPS are not available for Zambia. Nevertheless, inspection of Figure 17.6 shows price fluctuations of up to 20 percent within the nine-month period, so the impact of LRP is likely to pale to insignificance in comparison.

In conclusion, the price trends for beans and maize (accounting together for over 80 percent of project purchases) over the course of the nine procurement months are consistent with expected seasonal price trends. However, although the price fluctuations generated by the procurements of maize, vegetable oil, and HEPS would have fallen short of the magnitude of price variation from other effects, the opposite is true in the case of beans. There is a possibility that purchases of beans under the LRP project may have had a detectable market impact.

Retail Price for HEPS in Lusaka Central Market

8,000

9,000

10,000

11,000

12,000

13,000

14,000

14-D

ec-10

23-D

ec-10

25-Ja

n-11

22-F

eb-11

7-Mar-

11

22-M

ar-11

5-Apr-1

1

19-A

pr-11

5-May

-11

23-M

ay-11

8-Jun-1

1

28-Ju

n-11

13-Ju

l-11

26-Ju

l-11

8-Aug-1

1

22-A

ug-11

5-Sep

-11

Date

Pri

ce (Z

MK

/kg)

Retail Linear (Retail)


18. ZIMBABWE The United Methodist Committee on Relief (UMCOR) implemented an LRP project in Zimbabwe. The commodities procured for this project were all non-GMO, were purchased in Gauteng Province, South Africa and were distributed in Chipinge District, Zimbabwe. The commodities procured were 1,291 MT of maize grain, 233 MT of yellow peas, and 89 MT of Vitamin A-fortified cooking oil.

The procurement of white maize represents 0.2 percent of the white-maize production of 685,000 MT in 2009-2010 in Gauteng Province. This strongly suggests that, at the South African regional level, no market price impacts would have been likely in Gauteng Province as a result of this procurement. Since the national (South African) harvest was over 7.8 million MT, the relatively small LRP purchases would not have had any discernible price or supply impact at the national level.

However, as the UMCOR report notes:

In the current production season, 2010/2011, a projected drop in production of commercial maize has been estimated. More specifically, the Global Information and Early Warning System on Food and Agriculture (GIEWS) estimated that maize production will only reach 11.5 million MT in South Africa98 and that production will be 14 percent lower than last year. Heavy rains during the months of January and February also had a negative impact on yield of maize. The projected reduction in the crop through a smaller planting and poor weather has been pushing consumer and producer prices of white maize upwards.

This effect must be accounted for in any analysis of the price history of white maize from the 2010-2011 season purchased in South Africa. The historical trend in the price of maize in South Africa bears out these comments as seen in Figure 18.1, taken from the UMCOR report.

FIGURE 18.1: HISTORIC SOUTH AFRICAN WHITE MAIZE CONSUMER PRICES (RAND/MT)

Source: GIEWS, FAO, in USDA Local and Regional Procurement Pilot Project, Evaluation Report–Final LRP Project Report United Methodist Committee on Relief/Zimbabwe

98 We assume that the 11.5MT is a figure for white and yellow maize combined.


Figure 18.1 shows no spikes up or down in the price of white maize around the time of the procurement (marked by the vertical line). Rather, the prices show the steady but smooth upward trend noted by UMCOR in their report.

The cooking-oil situation is somewhat more complicated. Again from the UMCOR report,

Pricing information on cooking oil in South Africa is not publicly available.99 Hence, a proxy for sunflower oil prices derived from the SAFEX100 crude oil and oilcake prices was used to analyze price movements and trends. [A]s the comparison between monthly consumer prices for cooking oil and crude sunflower oil prices shows that change in consumer prices is fairly minor, the price of the crude product fluctuates more, indicating that UMCOR’s procurement was unlikely to affect consumer prices.

The data presented in Figure 18.2, also from the UMCOR report, support this assertion.

FIGURE 18.2: RELATIONSHIP BETWEEN SUNFLOWER (OIL & CAKE) SAFEX PRICES AND RETAIL PRICES

Source: SAFEX, in USDA Local and Regional Procurement Pilot Project, Evaluation Report–Final LRP Project Report United Methodist Committee on Relief/Zimbabwe

UMCOR purchased 89 MT of cooking oil, representing 0.037 percent of the national South African production of 242,805 MT of this commodity. Elasticity analysis suggests that such a relatively small purchase is highly unlikely to have caused any price spike, and the retail prices of the proxy commodity appear to be more stable than the export prices. However, the temporal resolution of the data is coarser than desired; weekly or bi-monthly time series would have been preferred for this analysis.

The UMCOR report notes that, historically, South Africa has produced relatively small amounts of peas, further noting that:

Since South Africa…imports the majority of its [bean] stock, impact of the procurement was evaluated based on the volume of imported peas available in the South African market. [Since] UMCOR purchased a mere 1.2 percent of the available dry pea stock in South Africa, there was no impact on the local demand and supply of dry peas to the South African market.

99 Pricing data were available for oilseeds and crude sunflower oil, which is sunflower oil priced prior to the capital intensive process of making the oil fit for human consumption. 100 SAFEX is the abbreviation for the South African Futures Exchange.


The data in table 18.1 support these assertions.

TABLE 18.1: SOUTH AFRICA DRY PEA PROCUREMENT EXPRESSED AS A SHARE OF DRY-PEA PRODUCTION, CONSUMPTION AND IMPORTS FROM CANADA

Metric Tons (MT) UMCOR Procurement Share

UMCOR procurement quantity 233 -

Total dry pea production 1,056 22.06%

Imports (from Canada) 18,155 1.28%

Domestic consumption - -

Total 19,211 1.21% Source: USDA Local and Regional Procurement Pilot Project, Evaluation Report–Final LRP Project Report United Methodist Committee on Relief/ZimbabweSource: UMCOR

Note that while the UMCOR procurement of peas represents 22 percent of the South African domestic production, the purchased amount (233 MT) represents only 1.21 percent of the total available in South Africa when the pea imports from Canada are added in to produce the total amount of this commodity available in South Africa.

The UMCOR report included limited time series data for the prices of maize and cooking oil. The commodity prices one week before procurement and one week after procurement were obtained from SAFEX. UMCOR explains this decision to use SAFEX prices by noting:

Spot prices on White Maize and Derived Sunflower (oil & cake) are directly obtained from SAFEX. Our report has made use of these prices because of the strong correlation of movement of spot prices together with commodity retail prices…. SAFEX prices minus the average transport differential and handling costs are the best estimates of farm gate prices.101

The evaluation team combined these limited data in Table 18.2, which shows the changes in prices over the two week period around the commodity purchases.

TABLE 18.2: TIME SERIES OF PRICES OF COMMODITIES PROCURED BY UMCOR

Commodity

Price 1 week Before Procurement

(South African Rand)

Price the Day of Procurement


Price 1 week After Procurement


Maximum Change from Price on the

Day of Procurement

White Maize 1,568 1,515 1,610 6.3%

Sunflower Oil and Cake 4,185.97 4,133.13 4,205.73 1.7%

Peas n/a 3,454.4 n/a n/a Source: SAFEX, in USDA Local and Regional Procurement Pilot Project, Evaluation Report–Final LRP Project Report United Methodist Committee on Relief/Zimbabwe.

The rise in the price of maize one week after the UMCOR procurement fits within the pattern of steadily rising maize prices during the four months after procurement described above was largely due to the steady price trends for maize. However, this does not explain why the price was lower on the day of the procurement than it had been one week earlier. Since this is the price that UMCOR actually paid and the others are SAFEX prices, the difference in data sources might explain this price difference.

101 National Agricultural Marketing Council (NAMC), Food Pricing Monitoring Committee report on “Analysis of selected food value chains”; December 2003


The price change in sunflower oil cannot be exactly tied to the price paid for the purchased commodity, since in this case the prices were for a proxy for the purchased commodity which, while a plausible substitute, may not bear a close enough similarity to it to stand up to close scrutiny of its price trend.

As noted earlier, there are too few data regarding yellow pea prices to create and assess even a limited time series of price data.

UMCOR reports that there were measurable impacts neither in the procurement market nor on their beneficiaries’ markets as a result of their project-related procurements. Although the data available were not sufficient to provide a more comprehensive independent analysis of potential market impacts, the evaluation team concludes that it is unlikely that the procurements of the listed commodities in South Africa had any discernible impact on the prices of these commodities there.


ANNEX IV. EVALUATION SCOPE OF WORK


CONTENTS

1. INTRODUCTION ........................................................................................................................... 209

2. BACKGROUND ............................................................................................................................. 209 2.1 Evaluation Audience .............................................................................................................. 210 2.2 References .............................................................................................................................. 210

3. SCOPE ............................................................................................................................................. 210

4. REQUIREMENTS .......................................................................................................................... 211 4.1 Phase I — Post-Award Meeting ............................................................................................ 212 4.2 Phase I — Final Work Plan ................................................................................................... 212 4.3 Phase I — Document Review ................................................................................................ 212 4.4 Phase 11— Domestic Interviews ........................................................................................... 212

4.4.1 Market Monitoring & The Learning Alliance .......................................................... 212 4.5 Phase 11— Pre-Departure Meeting & Field Work Plan ........................................................ 213 4.6 Phase III – Field Work ........................................................................................................... 213 4.7 Phase IV Post-Field Work Debrief and Report Outline ......................................................... 213 4.8 Phase V — Submit First Draft of Report ............................................................................... 214 4.9 Phase VI — First Draft Review Meeting ............................................................................... 214 4.10 Phase VII - Submit Second Draft of Report ........................................................................ 214

4.10.1 Second Draft Review Meeting (As Necessary) ..................................................... 214 4.10.2 Submit Third Draft of the Report ........................................................................... 214

4.11 Phase VIII — Submit Final Report ...................................................................................... 214

5. REPORTING .................................................................................................................................. 214 Quality Assurance Plan ................................................................................................................ 214 Monthly Performance Report ...................................................................................................... 215

6. DELIVERABLES ........................................................................................................................... 215

7. PERIOD OF PERFORMANCE .................................................................................................... 216

8. PLACE OF PERFORMANCE ...................................................................................................... 216

9. TRAVEL .......................................................................................................................................... 216

10. CONTRACT TYPE ........................................................................................................................ 216

11. STAFF QUALIFICATIONS .......................................................................................................... 216 Team Leader Capabilities ............................................................................................................ 217 Technical Specialist(s) Capabilities ............................................................................................. 217 Organizational Past Performance Records ................................................................................... 217

Appendix 1: USDA LRP Projects .......................................................................................................... 218

Appendix 2: USDA LRP Project Final Evaluation Requirements .................................................... 219

Appendix 3: Additional References ....................................................................................................... 221


1. INTRODUCTION The U.S. Department of Agriculture Local and Regional Food Aid Procurement Pilot Project (USDA LRP Project) is a five-year, $60 million pilot authorized in Section 3206 of the Food, Conservation, and Energy Act of 2008 (Farm Bill) for the purpose of examining the timeliness and efficiency of using local and regional procurement (LRP) as a tool to enhance U.S. Government food assistance programs. The rationale for LRP is based on the premise that providing cash grants for the purchase of food from surplus markets in the recipient country or region may allow for a quicker and more cost-efficient response to a natural disaster or other food crisis.

Although the project’s focus is on funding emergency programs, the legislation directed USDA to use a portion of the funds for field-based projects that also provide development assistance. However, development programs were required to be for a period of no less than one year, which in effect prohibited the funding of development programs after October 1, 2010. In addition, the legislation directed USDA to select a diversity of projects, including projects located in multiple geographic regions as well as both food surplus and food deficit regions. However, the legislation also required that USDA ensure that the majority of the field-based projects selected be located in Africa and procure eligible commodities produced in Africa.

The USDA LRP Project is being managed by the Food Assistance Division (FAD) of the Office of Capacity Building and Development (OCBD), Foreign Agricultural Service (FAS), of the U.S. Department of Agriculture (USDA). In accordance with the requirements of the authorizing law, USDA seeks to hire an external contractor to conduct an independent evaluation of the USDA LRP Project.

2. BACKGROUND Under the 2008 Farm Bill, Congress authorized $60 million in Commodity Credit Corporation funding for USDA to implement a local and regional food aid procurement pilot project in FY 2009 – 2012. The pilot is intended to study the timeliness and effectiveness of using local and regional procurement of food aid in responding to food crises and disasters. The authorizing law required USDA to do the following:

1) Complete and submit a report to Congress no later than 180 days after the date of enactment that studied:

o Local and regional procurements for food aid programs conducted by- other donor countries; private voluntary organizations; and The World Food Programme (WFP) of the United Nations.

2) Issue guidelines to carry out field-based projects that take into account the results of the study. 3) Provide grants to, or enter into cooperative agreements with, eligible organizations to carry out

field-based projects that consist of local or regional procurements of eligible commodities to respond to food crises and disasters.

4) Submit to the appropriate committees of Congress a report that contains the analysis and findings of an independent evaluation no later than four years after the date of enactment (June 17, 2012).

In fiscal year 2009, USDA fulfilled the first requirement of the authorizing law by completing the required report, The Use of Local and Regional Procurement in Meeting the Food Needs of Those Affected by Disasters and Food Crises.102 This study was submitted to the appropriate committees of Congress on January 15, 2009. USDA fulfilled the second requirement by issuing Interim Guidelines for the Local and Regional Food Aid Procurement Pilot Project103 on September 21, 2009. From fiscal year 2009 through 2011, USDA fulfilled the third requirement of the authorizing law by providing funding for the implementation of 21 field-based projects with seven Private 102http://www.fas.usda.gov/excrodits/FoodAid/LRP/USDALRPStudy.pdf 103http://www.fas,usda.gov/excredits/FoodAid/LRP/Interim PPP Guidelines.pdf

http://www.fas.usda.gov/excrodits/FoodAid/LRP/USDALRPStudy.pdf

http://www.fas/

http://usda.gov/excredits/FoodAid/LRP/Interim


Voluntary Organizations (PVOs) and WFP to implement emergency response and development food assistance programs in Sub-Saharan Africa, Central America, and South & Southeast Asia (see Appendix 1 for a detailed listing of the projects and grant recipients).

This request for quote (RFQ) is being issued to fulfill the fourth requirement under the authorizing law, the submission to the appropriate committees of Congress a report that contains the analysis and findings of an independent evaluation (see Appendix 2 for the detailed requirements of the evaluation as specified in the authorizing law).

2.1 Evaluation Audience The main audience for the independent evaluation of the USDA LRP Project report is the appropriate committees of the U.S. Congress. More specifically, the report will be reviewed by the Senate Committee on Agriculture, Nutrition and Forestry, and House of Representatives’ Committees on Agriculture and Foreign Affairs. Within USDA, the report will be widely circulated, with special attention paid by FAS’ overseas posts. Externally, other U.S. governmental agencies such as the Department of State and the U.S. Agency for International Development (USAID), the PVO community, WFP, universities, foreign governments and the international food security audience writ large will be interested in the findings of this evaluation.

2.2 References The following references are attached as part of this requirement for the contractor’s review:

The Use of Local and Regional Procurement in Meeting the Food Needs of Those Affected by Disasters and Food Crises104

INTERNATIONAL FOOD ASSISTANCE: Local and Regional Procurement Can Enhance the Efficiency of U.S. Food Aid, but Challenges May Constrain Its Implementation105

Market Information and Food Insecurity Response Analysis. Barrett, Bell, Lentz and Maxwell. February 2009.106

Section 3206 of the Food, Conservation, and Energy Act of 2008107 Interim Guidelines for the Local and Regional Food Aid Procurement Pilot Project108 See Appendix 3 for Additional References

3. SCOPE The contractor is expected to perform the requirements of this contract in two periods as identified in Section 4 below. In addition to the tasks summarized below, within one week of the contract award, the contractor will be required to electronically submit a Quality Assurance Plan to the Contracting Officer’s Technical Representative (COTR).

In general, the selected contractor will perform the following key activities in the first period:

Review relevant background documentation including, but not limited to, the documents referenced in Section 2.2, PVO and WFP proposals, agreements and other relevant materials. FAD will provide the Contractor with copies of the relevant proposals and agreements.

Review PVO quarterly and cumulative procurement and distribution status reports, other programmatic reports, commodity quality inspection reports, and other materials that become available. FAD will provide the Contractor with copies of these documents.

104http://www.fas.usda.gov/excredits/FoodAici/LRP/USDALRPStudy.pdf 105http://www.gao.gov/newitems/d09570.pdf 106http://www.basiswisc.edu/ept/barrett%20background%20food%20security.pdf 107http://www.fas.usda.gov/excredits/FoodAid/LRP/Farm_Bill_2008.pdf 108http://www.fas.usda.gov/excredits/FoodAid/LRP/Interim_ PPP_Guidelines.pdf

http://www.fas.usda.gov/excredits/FoodAici/LRP/USDALRPStudy.pdf

http://www.fas.usda.gov/excredits/FoodAid/LRP/Farm_Bill_2008.pdf

http://www.fas.usda.gov/excredits/FoodAid/LRP/Interim


Meet with Chris Barrett and Erin Lentz from Cornell University to gain a comprehensive understanding of their work and the methodology that the Learning Alliance members are employing.

Meet with WFP to gain a comprehensive understanding of the market monitoring strategies being employed under the Purchase for Progress initiative and other operations that are conducting local and regional procurements (Emergency Operations {EMOPs}, Protracted Relief and Recovery Operations {PRROs}and Development Operations {DEVs}).

Meet with the authors of the GAO report to gain a comprehensive understanding of the methodology employed to develop their report and obtain lessons learned from the process.

Develop a methodology for conducting the evaluation and statistical analysis that fulfills the requirements set out in the authorizing law.

Develop a work plan to implement the evaluation methodology. Visit a sampling of USDA LRP Project field-based projects in each geographic region to collect

primary data, including, but not limited to: first-person interviews with PVO and WFP staff, project beneficiaries, participating vendors, project sub-recipients (particularly those involved in the market monitoring component), warehouses and other relevant project stakeholders

In general, the selected contractor will perform the following key activities during option period 1:

Review final market analysis reports submitted by participants, including reports from project price monitoring databases and other data collected by participants. FAD will provide the Contractor with copies of these documents.

Provide FAD with the initial draft of the report. o The report should utilize and build upon the aforementioned GAO report. o FAD will assist the Contractor to access data comparable to that used in the GAO report as

well as that collected by the PVOs that received funding under the USDA LRP Project. Respond to comments and questions from FAD and revise the report as necessary. Be available to respond to a second round of questions and comments from FAD, as necessary. Finalize the report for the appropriate committees of the U.S. Congress.

4. REQUIREMENTS The Contractor shall facilitate and guide an independent evaluation of all field-based projects implemented under the USDA LRP Project in accordance with P.L. 110, Section 3206 (f) of the authorizing law. The primary purpose of the final evaluation is to address each of the factors listed in P.L. 110, Section 3206, part (f). The contractor shall also include a separate analysis utilizing a sample of countries to examine whether local and regional procurement is timelier (when considering both the speed of procurement and delivery) and more cost-effective than in-kind food aid.

As part of this evaluation, the Contractor is expected to review existing literature, obtain and analyze relevant data, interview relevant stakeholders and other primary sources and conduct an overall analysis of the USDA LRP Project based on the requirements of the authorizing law. Prior to initiating the evaluation, the Contractor will present a detailed implementation plan including the assessment methodology and timeline for conducting the work and how the evaluation findings will be presented to FAD. The Contractor is expected to complete a comprehensive evaluation report that fulfills the requirements denoted in the 2008 Farm Bill (P.L. 110, Section 3206, Part (f)) to be submitted to appropriate committees of the Congress and present the findings in a public information session.

Throughout the course of the evaluation, the Contractor’s participation shall be limited to the activity for which the Contractor is responsible under the statement of work. The Contractor’s employees shall identify themselves as contractor personnel when in meetings or engaging in activities associated with the requirements in the statement of work.


FIRST PERIOD

4.1 Phase I — Post-Award Meeting Within one week following the contract award, the Contractor will work with the LRP Project Manager to arrange to meet with FAD, in particular the USDA LRP Project staff, to discuss the proposed evaluation methodology. The purpose of this meeting is to provide the Contractor and FAD with an opportunity to obtain mutual clarification regarding the requirements of the statement of work and the proposed evaluation methodology respectively. This meeting will also serve as an opportunity to discuss potential revisions to the evaluation methodology and work plan, as necessary.

4.2 Phase I — Final Work Plan Within one week of the post-award meeting, the contractor will electronically submit to the COTR and LRP Project Manager a finalized detailed work plan that incorporates feedback from the post-award meeting and includes a chronology of how the requirements in this statement of work will be met. The Contractor will not begin Phase II until it receives written acceptance of the work plan from the LRP Project Manager.

4.3 Phase I — Document Review The Contractor will gather and study relevant documentation (including, but not limited to, the materials referenced in Sections 2.2 and 3) and existing data sets to obtain a comprehensive understanding of the history of local and regional procurement and accepted methodologies, for project implementation. FAD will provide the Contractor with documents as described in Section 3, and will assist the Contractor to obtain the relevant data sets, as necessary.

4.4 Phase 11— Domestic Interviews In addition to reviewing the relevant documentation, the Contractor will meet with stakeholders in the U.S., including, but not limited to: relevant staff at Cornell University, participating PV0s and WFP, as well as the authors of the GAO report. The Contractor will also study the Learning Alliance market monitoring strategy, Learning Alliance market prices database, and the dataset used to develop the GAO report.

4.4.1 Market Monitoring & The Learning Alliance

The 2008 Farm Bill outlined market monitoring requirements to ensure that the local and regional procurements did not have a disruptive impact on markets. Specifically, the legislation directed USDA to ensure that locally and regionally-procured food aid was provided to affected populations without significantly increasing commodity costs for low-income consumers who procure commodities from the same markets where the eligible commodities were procured: In addition, USDA was directed to ensure that the procurement and distribution of the commodities did not have a disruptive impact on farmers located in, or the economy of, the recipient country or any country in the region in which the commodities were procured and/or distributed. The legislation also directed USDA to take precautions to ensure that the procurement of the eligible commodities did not unduly disrupt world prices for agricultural commodities or normal patterns of commercial trade with foreign countries, and that the procurements were made at a reasonable market price with respect to the economy of the country in which the eligible commodities were procured.

To ensure the fulfillment of these requirements, six of the seven PVO grantees created a consortium called The Learning Alliance. The Learning Alliance, which includes Cornell University as the technical lead, is a mechanism for the PVOs to develop a standardized strategy for the implementation of the market impact monitoring component of their respective field-based projects. The Learning Alliance members are responsible for implementing 13 of the 14 field-based projects awarded to PVOs under the USDA LRP Project. The Learning Alliance members have also pooled their monitoring and evaluation funds to create an


online database in which they record the market prices they have collected throughout the program implementation period.

In addition to leading the development of the Learning Alliance’s market impact monitoring strategy, the Cornell University staff, led by Chris Barrett, are assisting PVOs with their pre-purchase analyses, the collection and analysis of secondary data and are providing ongoing technical support.

Cornell University is also conducting in-depth evaluations with counterfactuals for three field-based projects (CRS - Burkina Faso, CRS — Guatemala and Land O’Lakes - Zambia). These in-depth project evaluations will utilize matched communities and in-kind food aid projects as benchmarks against which the market impacts and effectiveness, of local and regional procurements can be more effectively assessed. Thus, the work conducted by Cornell University under the Learning Alliance is an integral component of the USDA LRP Project that the Contractor is expected to assess comprehensively.

WFP and PVO(s) not involved in the Learning Alliance are employing other approaches to market monitoring and are ensuring that their procurements do not result in disruptive market impacts. The Contractor is expected to also examine the methodologies employed by these organizations.

4.5 Phase 11— Pre-Departure Meeting & Field Work Plan Prior to conducting the field work, and no later than July 15, 2011, the Contractor will contact the LRP Project Manager to schedule a meeting with FAD to provide feedback from the domestic interviews and discuss a detailed work plan for the upcoming field work. At least three business days prior to the meeting, the Contractor will electronically submit the detailed field work plan to the COTR and the LRP Project Manager. The Contractor will not begin Phase III until written acceptance of the field work plan from the LRP Project Manager is received.

4.6 Phase III – Field Work Following the approval of the field work plan, but not later than July 30, 2011, the Contractor will begin travel to the agreed upon USDA LRP Project field-based projects to conduct primary research for the evaluation. The field work must be concluded by September 30, 2011. The Contractor will meet with LRP Project stakeholders, including but not limited to: PVO and WFP staff, external organizations involved in monitoring the projects’ impacts on markets, beneficiaries, participating vendors, grantee sub-recipients, warehouse staff, and other relevant project stakeholders. The Contractor will review relevant documentation from the field-based projects including, but not limited to: proposals, project implementation plans, procurement files, market monitoring reports, financial documents related to procurements and voucher activities, food quality inspection reports, and warehouse inventory reports. The Contractor will collect additional commodity price data as necessary. The Contractor will also review all available PVO quarterly and cumulative procurement and distribution reports, end of project implementation and market analysis reports/data.

SECOND PERIOD

4.7 Phase IV Post-Field Work Debrief and Report Outline Within two weeks of the completion of Phase III, the Contractor will contact the LRP Project Manager to schedule a meeting with FAD. The meeting must take place no later than October 17, 2011. At least five work days prior to the meeting the Contractor will electronically submit an outline of the evaluation report to the COTR and the LRP Project Manager. At the meeting, the Contractor and FAD will discuss the findings from the field work and outline of the evaluation report to ensure that the Contractor’s work aligns with the requirements of the authorizing law prior to beginning to draft the report. The Contractor will not begin Phase V until it receives written acceptance of the outline of the evaluation report from the LRP Project Manager.


4.8 Phase V — Submit First Draft of Report No later than March 15, 2012 the Contractor will electronically submit the first full draft of the evaluation report to the COTR and LRP Project Manager.

4.9 Phase VI — First Draft Review Meeting Within 7 days of the submission of the first draft of the report, the Contractor and LRP Project Manager will schedule a Draft Review Meeting. At this meeting, FAD will pose questions and provide feedback to the Contractor regarding the submitted report. At the end of the meeting the Contractor will provide a list of outstanding issues and/or revisions that will be addressed in the subsequent version of the report.

4.10 Phase VII - Submit Second Draft of Report No later than April 11, 2012, the Contractor will submit a revised draft of the report that addresses all of the comments and questions from the Draft Review Meeting, to the COTR and LRP Project Manager.

4.10.1 Second Draft Review Meeting (As Necessary)

If deemed necessary by FAD, the Contractor and LRP Project Manager will schedule a Second Draft Review Meeting to occur no later than April 27, 2012. At this meeting, FAD will pose questions and provide feedback to the Contractor regarding the submitted report. At the end of the meeting the Contractor will provide a list of outstanding issues and/or revisions that will be addressed in the subsequent version of the report.

4.10.2 Submit Third Draft of the Report

No later than May 4, 2012, the Contractor will submit a revised draft of the report that addresses all of the comments and questions from the Second Draft Review Meeting.

4.11 Phase VIII — Submit Final Report The Contractor will electronically submit a copy of the final report to the COTR no later than July 24, 2012. The Contract shall also submit 125 bound hard copies of the final report printed in color and professionally finished for approval to the LRP Project Manager no later than December 28, 2012. The Contractor will receive written approval of the report from the LRP Project Manager.

5. REPORTING The Contractor shall provide all of the reports below to enable the LRP Project Manager and COTR to monitor progress and ensure compliance. All reports prepared by the Contractor shall be submitted to the LRP Project Manager and COTR for review and approval.

Quality Assurance Plan The Contractor shall provide a written Quality Assurance Plan that defines the activities, roles and responsibilities planned to ensure successful completion and performance in accordance with the requirements. The plan should identify how the Contractor will ensure the timeliness and acceptability of deliverables; monitor and track productivity and resolve problems in a timely manner. This report shall be submitted electronically to the COTR within one week of the contract award.


Monthly Performance Report The Contractor shall provide a written monthly report that tracks progress toward completion of tasks identified throughout section 4 above and deliverables, identified in the table in section 6 below. The monthly report will be due the last working day of each month and should be a maximum of 3 pages. In addition to the narrative update, the monthly report should include a table that lists the tasks and deliverables in this requirement and their completion dates. The Contractor’s final monthly report shall be a closeout report.

6. DELIVERABLES The Contractor shall provide the items listed in the table below to the USDA LRP Project staff and COTR throughout the contract’s period of performance. The table provides the due date and phase each deliverable will occur. The table also estimates the level of effort that may be expended on each deliverable as expressed as a percentage.

BASE PERIOD

Deliverable Description SOW Reference

Estimated Deliverable Due Date

Unit of Issue

Quality Assurance Plan Section 5 Within I week of contract award (1) one Monthly Performance Reports Section 5 Last working day of each month (3) three

Post Award Meeting Section 4.1 Week of July 4th (no later than July 8, 2011) N/A

Submit Final Work Plan Section 4.2.1 Within 1 week of post award meeting (1) one

Submit Field Work Plan Section 4.5 Three Days Prior to Pre-Departure Meeting (1) one

Pre-Departure Meeting Section 4.5 Week of July 18, 2011 (1) one

OPTION PERIOD

Deliverable Description SOW Reference

Estimated Deliverable Due Date

Unit of Issue

Monthly Performance Reports Section 5 Last working day of each month (14) fourteen

Submit Outline of Evaluation Report Section 4.7 Five Days Prior to Post-Field Work

Debrief Meeting (1) one

Post-Field Work Debrief and Report Outline Review Meeting Section 4.7 Week of October 17, 2011 (1) one

Submit First Draft of Report Section 4.8 March 15, 2012 (1) one First Draft Review Meeting Section 4.9 Week of March 19, 2012 (1) one Submit Second Draft of Report Section 4.10 April 11, 2012 (1) one USDA Comments on Second Draft Received by MSI Section 4.10 April 25, 2012 (1) one

As Necessary: Second Draft Review Meeting Section 4.10.1 No Later than April 27, 2012 (1) one

Submit Third Draft of Report Section 4.10.2 May 4, 2012 (1) one


USDA Comments on Third Draft Received by MSI Section 4.10.2 June 8, 2012 (1) one

Submit Revised Third Draft of the Report (submitted in sections) Section 4.10.2 June 20, 2012 –July 18, 2012 (1) one

USDA Comments on Third Draft Received by MSI Section 4.10.2 June 27, 2012 – July 20, 2012 (1) one

Submit Final Report Section 4.11 July 24, 2012 (1) one

7. PERIOD OF PERFORMANCE The period of performance for this requirement will begin on the award date and end no later than December 28, 2012.

8. PLACE OF PERFORMANCE The place of performance for this contract will be the Washington D.C. Metro Area, except for travel as specified in Section 9. All work will be performed at the Contractor’s facilities. The Contractor will be responsible for making all arrangements to visit with and interview relevant stakeholders. Likewise, the Contractor will be responsible for making all travel arrangements to visit the field-based projects and for obtaining necessary work space or other accommodations.

9. TRAVEL The Contractor is expected to visit a number of the field-based projects funded under the USDA LRP Project to conduct primary research. The Contractor is also expected to travel to Ithaca, NY to meet with the relevant Cornell University staff. The Contractor should denote each planned trip in the proposal, including: the countries to be visited, number of days per country and number of staff per trip (see Appendix 1 for a detailed listing of the projects and grant recipients).

10. CONTRACT TYPE This is a fixed price contract.

11. STAFF QUALIFICATIONS The Contractor shall provide staff to perform listed requirements within the statement of work in whatever mix of labor categories and whatever quantity of personnel/man-hours it determines to be appropriate to optimally meet the requirements stated herein. The mandatory specific personnel labor required are the Team Leader and Technical Specialist(s) (see below). These positions are designated as Key Personnel. The individuals proposed for these Key Personnel positions are considered by FAD to be essential to task order performance. The Contractor must re-confirm the availability of the individuals proposed for the Key Personnel positions, for the duration of the task order, prior to the execution of the contract. The Contracting Officer must be notified in writing at least 60 days prior to the Contractor removing or changing any of the Key Personnel from the order.

The Contractor shall submit curriculum vitae for each Key Personnel. The contractor shall demonstrate that each Key Personnel proposed possesses the required skills and experience. The Contractor shall detail the percentage of each Key Personnel staff member’s time that shall be devoted to this project versus other


projects throughout the duration of the evaluation. After award, any changes to the Key Personnel must be approved by the COTR and LRP Project Manager. Any proposed replacement must possess similar and comparable skills and expertise when compared to the original proposed Key Person. Hence, all Key Personnel curricula vitae must be reviewed and approved prior to new Key Personnel beginning work under this contract.

Team Leader Capabilities The Contractor shall ensure that one of the key personnel will be the Team Leader who will bear the overall responsibility for the evaluation. The Team Leader shall be a Senior-level person with at least 15 years of experience leading evaluation teams to conduct complex economic analyses in developing countries, including primary research. The Team Leader must possess a history of successfully completing high-profile evaluations for the U.S. government. Previous experience completing comparable work for USAID and/or previous experience writing Congressional reports is preferred. The Team Leader is expected to have a strong background in multiple facets of food assistance programming, preferably from the perspective of both an implementing organization and a donor. The Team Leader should have a demonstrated capacity for conducting original econometric analyses. The Team Leader is expected to have significant experience living/working overseas in multiple geographic regions, including Sub-Saharan Africa. The Team Leader should also have strong communication (spoken, written and presentational), management and leadership skills including the ability to defend or explain difficult issues or positions. The quote must include the Team Leader’s detailed CV, including a reference (preferably a supervisor and/or donor representative) for each position held for a minimum of the past five years or five positions (whichever is greater). In addition, a writing sample for which the Team Leader was the primary author must be submitted with the application.

Technical Specialist(s) Capabilities The Technical Specialist(s) will assist the Team Leader in the design, analysis and writing of the project evaluation. The Technical Specialist(s) shall be a Mid/Senior-level person with at least five years of experience working on evaluation teams focused on conducting complex economic analyses in developing countries, including collecting primary research. The Technical Specialist(s) must possess a history of participating in successfully completed high-profile evaluations for the U.S. government. Previous experience working on comparable evaluations for USAID and/or previous experience writing Congressional reports is preferred. The Technical Specialist(s) is expected to have a strong background in food assistance programming, preferably from the perspective of both an implementing organization and a donor. The Technical Specialist(s) should have a demonstrated capacity for conducting original econometric analyses. The Technical Specialist(s) is expected to have significant experience living/working overseas in multiple geographic regions. The Technical Specialist(s) should also have strong communication (spoken, written and presentational) skills including the ability to, defend or explain difficult issues or positions. The application must include the Technical Specialist’(s)’ detailed CV, including a reference (preferably a supervisor and/or donor representative) for each position held for a minimum of the past five years or five positions (whichever is greater). In addition, a writing sample for which the Technical Specialist was the primary author must be submitted with the application.

Organizational Past Performance Records In addition to the capabilities of Key Personnel, the Contractor must provide details of their past performance on previously-executed contracts to conduct independent evaluations of government projects. In particular, the Contractor should provide details on the organization’s previous experience conducting evaluations of food assistance programming, food security-related issues, other international development projects, as well as market analyses and econometric analyses. For each previously-executed contract, the Contractor must provide contact details for the Contracting Office or other client-side point of contact.


Appendix 1: USDA LRP Projects109

Fiscal Year Organization Recipient Country Program Type

1 2009 WFP Mali Development

2 2009 WFP Malawi Development

3 2009 WFP Tanzania Development

4 2010 Mercy Corps Niger Emergency

5 2010 Land O’Lakes Bangladesh Development

6 2010 Catholic Relief Services (CRS) Guatemala Emergency

7 2010 CRS Mali Development

8 2010 CRS Benin Development

9 2010 CRS Burkina Faso Development

10 2010 International Relief and Development Cambodia Development

11 2010 Land O’Lakes Zambia Development

12 2010 WFP Mali Development

13 2010 Fabretto Children’s Foundation Nicaragua Development

14 2010 WFP Congo, Republic of Emergency

15 2010 WFP Cameroon Emergency

16 2010. WFP Chad Emergency

17 2011 World Vision Uganda Development

18 2011 WFP Cameroon Emergency

19 2011 United Methodist Committee on Relief Zimbabwe Emergency

20 2011 World Vision Kenya Emergency

21 2011 CRS Niger Emergency

22 2011 WFP Mozambique Emergency

23 2011 WFP Pakistan Emergency

109 This list includes 23 projects, per the original SOW. However, two WFP-implemented LRP projects were funded for two fiscal years each, but in each case, this evaluation treats the project individually, in accordance with WFP’s project reports. One project (in Benin) was ended before procurement began. The net total, then, is 20 projects in 18 countries, and those figures are used throughout this report.


Appendix 2: USDA LRP Project Final Evaluation Requirements (Due to Congress by June 17, 2012) The purpose of the final evaluation is to address each of the factors listed in P.L 110, Section 3206, Part (f). The Contractor shall also include a separate analysis utilizing a sample of countries to examine whether local and regional procurement is timelier (when considering both the speed of procurement and delivery) and more cost-effective than in-kind food aid.

The evaluator will have to take into consideration whether a program is for emergency or development assistance, as the authorizing law requires that USDA provide funding for both types of programs. The expeditious procurement and delivery of food assistance during an emergency is especially critical. The speed of the procurement and delivery is also important for development programs, but in general, this element is less critical as development programs tend to address chronic food insecurity in situations in which lives are not immediately at stake. Cost is an important element to examine for both emergency and development assistance programs.

A list of each of the specific factors that must be addressed in accordance with P.L. 110, Section 3206, Part (f) is provided below. Each of these factors must be addressed in the body of the final evaluation report.

i. With respect to each relevant market in which an eligible commodity was procured, a description of:

1. The prevailing and historic supply, demand and price, movements of the market (including the extent of competition for procurement bids);

2. The impact of the procurement of the eligible commodity on producer and consumer prices in the market;

3. Each government market interference or other activity of the donor country that might have significantly affected the supply or demand of the eligible commodity in the area at which the local or regional procurement occurred;

4. The quantities and types of eligible commodities procured in the market; 5. The timeframe for the procurement of each eligible commodity; 6. The total cost of the procurement of each eligible commodity (including storage,

handling, transportation and administrative costs); 7. An assessment regarding whether the requirements of 3206(e) and (f) have been met; 8. The impact of different methodologies and approaches on:

a. Local and regional agricultural producers (including large and small agricultural producers);

b. Markets; c. Low-income consumers; and d. Program recipients; and

9. The length of the period beginning on the date on which the Secretary initiated the procurement process and ending on the date of delivery of the commodities;

10. A comparison of different methodologies used...with respect to: a. The benefits to local agriculture; b. The impact on markets and consumers; c. The period of time required for procurement and delivery; d. Quality and safety assurances; and e. Implementation costs; and


11. To the extent to which adequate information is available, including the results of the initial study conducted under the USDA LRP Project, a comparison of the different methodologies used by other donor countries to make local and regional procurements.


Appendix 3: Additional References A Market Analysis and Decision Tree Tool for Response Analysis: Cash, Local Purchase and/or Imported Food Aid? Daniel G. Maxwell, Erin C. Lentz, and Christopher B. Barrett, Cooperative for Assistance and Relief Everywhere (CARE), May 2007. http://aem.cornell.edu/faculty_sites/cbb2/Papers/Decision%20Tree%20Tool%20%20May%2023%202007.pdf Local and Regional Food Aid Procurement: An Assessment of Experience in Africa and Elements of Good Donor Practice, David Tschirley, and Anne Marie Del Castillo, Michigan State University International Development Working Paper No. 91, 2007.

Year: 2001-2005 Country: Kenya, Uganda, Zambia and Mozambique Commodity: Corn and corn/soy blend

The United States’ International Food Assistance Programs: Issues and Options for the 2007 Farm Bill, Christopher B. Barrett, February 2007.

Year: 2007 Country: United States Commodity: Not applicable; general discussion of U.S. food aid

The Development Effectiveness of Food Aid: Does Tying Matter? Organization for Economic Cooperation and Development, 2006

Year: 2002-2003 Country: Various donating and recipient countries Commodity: Wheat, corn, corn-soy blend, vegetable oil, and rice

World Food Program Studies/Reports: Food Procurement in Developing Countries, World Food Program, Executive Board, First Regular Session, Feb, 2006, Rome.

Report summarizes other WFP studies on LRP in Bolivia, Burkina Faso, Ethiopia, Nepal, South Africa, Uganda, and Congo

World Food Program Local and Regional Food Procurement-An Analytical Review (Ethiopian Case Study), Final Report. Addis Ababa: June 2005.

Year: 2001-2004 Country: Ethiopia Commodity: Corn and wheat

Local and Regional Food Procurement in Uganda an Analytical Review, A study report prepared for the Economic Analysis and Development Policy Unit in the Strategy, Policy and Program Support Division of the World Food Program, Serunkuuma and Associates Consult, June 2005.

Year: 2001-2004 Country: Uganda Commodity: Corn and beans

Food Aid Procurement in South Africa: an Analytical Review of WFP Activities; Nick Vink, Thulasizwe Mkhabela, Ferdie Meyer, and Johann Kirsten; April 2005.

Year: 2001-2004 Country: South Africa Commodity: Corn


Democratic Republic of Congo Food Procurement Assessment Mission Euateur, Katanga, Orientale, North Kivu and South Kivu Provinces; World Food Program; May 2007.

Years: 2001-2006 Country: Democratic Republic of Congo Commodity: Corn and pulses

Impact of WFP’s Local and Regional Food Purchases (A Study Case on Burkina Faso) Final Report Submitted by Institut de Sahel Comite’ Permanent Inter-Etats de Lutte Contre La Secheresse dans le Sahel, Mali.

Year: 2002-2005 Country: Burkina Faso Commodity: Corn, corn meal, sorghum, and cowpea


ANNEX V. DATA COLLECTION INSTRUMENTS


CONTENTS

1. COUNTRY CHECKLIST ................................................................................................. 225

2. INTERVIEW QUESTIONS FOR PVO REPRESENTATIVES .................................... 228

3. BENEFICIARY INTERVIEW (CUSTOMIZED BY COUNTRY; CAMBODIA IRD PROJECT VERSION INCLUDED HERE, OTHERS AVAILABLE ON REQUEST) ........................................................................................ 233

4. NON-BENEFICIARY INTERVIEW (AS ABOVE; ZAMBIA LAND O’LAKES PROJECT VERSION INCLUDED HERE) ................................................. 236

5. WHOLESALE VENDOR SURVEY INSTRUMENT ..................................................... 237

6. VENDOR QUESTIONNAIRE – RETAIL (AS ABOVE; UGANDA WORLD VISION PROJECT VERSION INCLUDED HERE) .................................................... 239

Instruments were customized by country and project type, particularly for beneficiaries, non-beneficiaries and local vendors (in the case of voucher programs). All interview instruments served as guides but allowed for customization while in the field or on the phone with respondents, in order to understand the variability of LRP projects.

The PVO instrument was used for U.S. and in-country interviews, though for each use, respondents were unable to answer all questions. By attempting to do the full interview with each, the evaluation team’s interviews opened up additional lines of inquiry when answers did not correspond.


1. Country Checklist

Instrument/tool Timeliness/

Cost-effectiveness comparisons

Market impact Quality/suitability/ acceptability Other

LRP PVO Interview(s)

• Develop procurement timeline from start-up (including pre-auth)to completion of the first round of distributions

• Get accounting of expenditures on main categories - labor, transport, procurement, admin, indirects, etc. (totaling all expenses, including any that fall outside budget)

• IPP analysis methodology for LRP and the non-commodity price data and technical coefficients necessary to implement it

• Time series on market prices – get actual data files

• Opinions, experiences of market impacts

• Market monitoring report

• Market conditions • Market dynamics and

structure [What do “dynamics and structure” mean to our team?]

• Evidence of vendor participation/collusion, sales of commodities

• Impacts on farmers, esp. capacity building projects

• Commodity selection process

• Testing process, results, issues

• PVO opinions

• Project description, procurement modalities, team make-up

• Process questions from instrument – contracts, transport, staffing, process comparison with in-kind aid, decision processes, monitoring progress (M&E), impacts, successes, lessons learned

• Delays – communication, transport (road/river conditions, blocks, state and availability of trucks/ boats, drivers and labor, admin delays, etc), testing, USDA contract / amendments

PVOs involved in in-kind food aid (may be the LRP PVO, WFP or other)

• Garner agreement to provide data on costs and timeline for in-kind aid after it reaches port and out to distribution node

• Primary, secondary market price data to make decisions, inform program

• Market conditions • Market dynamics,

structure

• Testing procedures • Lab quality/stringency

estimates • PVO opinions

• Evidence of shocks or large gov’t purchases, other events or conditions that may have affected market

Vendors (wholesale) • Times to complete procurement and shipment

• Pricing of commodities

• Market estimates, relative size of LRP

• Estimates of market impact of LRP

• Testing process • Frequency of and

reasons for losses during transport

• Bid/tender process • Process questions –

contracts, logistics, benefits to suppliers or producers





Vendors (retail)

• [These are small sellers near beneficiaries]

• Market disruptions, impact

• Supply disruptions • Re-selling of goods

• Quality changes • Process questions – such as voucher issues

Farmers (as applicable)

• Market disruptions, price changes

• Farm-gate prices, prices received from wholesale vendor(s), modality of price setting (pre-harvest agreements, market price

• Supply disruptions

Beneficiaries • Timeliness and need • Market disruptions, price changes

• Supply disruptions • Selling and what was

purchased w/proceeds

• Quality, preferences • Program impacts, delivery and transport

• Coping strategies

Non-beneficiaries

• Market disruptions, price changes

• Supply disruptions • Re-selling of goods

• Coping strategies

Labs/testing agencies

• Turnaround time • Other bottlenecks

• Standards • Sampling methods • Outsourcing • Rejected orders • Comparison with other

food aid orgs

USAID/USDA (as applicable)

• FEWSNET or other data sources

Public sector • Estimates of need and seasonality of need

• Secondary data on prices

• Testing facilities and quality in-country

• Standards





Other subcontracts (warehouse, transport, Customs (or other handling/clearing)

• IPP or other transport costs

• Timelines and bottlenecks

• Quality


2. Interview Questions for PVO Representatives COUNTRY

ORGANIZATION

RESPONDENT(S)/ ROLE(S)

U.S. or IN-COUNTRY IW EMERGENCY/DEVELOPMENT

PROJECT BUDGET

1. Please tell us about your LRP pilot project(s): a. Which commodity(ies) was/were procured? How were these selected, and what were your

criteria? b. What was the procurement modality (i.e. competitive/non-competitive contract, spot purchase,

voucher, etc.)? c. Who conducted the procurement? (PVO’s in-house procurement staff, or a hired agent to

manage the procurement process?) d. Were there any local laws that impacted the procurement timeline (i.e. mandatory requirements

to advertise the invitation for bids for a specific length of time, etc?) e. Who were the commodity vendors? f. How many vendors supplied commodities for the project? g. What was the extent of competition in the market? (i.e. the overall number of vendors in the

market, the number of qualified vendors responding to invitations for bids, etc.) h. Was there any evidence of vendor collusion and price-setting? i. Time period of project and frequency of procurement? (Were there multiple invitations for bid,

long-term contracts, etc?) j. Sites for procurement and distribution (Were the commodities purchased in retail markets,

wholesale markets, from small-holder farmer cooperatives, from commodity exchanges, etc?) k. How were procurement sites chosen? (Did the PVO understand the market dynamics and have

a methodology for identifying a market surplus? Did they continuously monitor market supply to make certain that the procurement would have no negative impact on low-income consumers?)

l. In what season were commodities procured (i.e. harvest season, lean season, etc.)? (Note: If there was verification of adequate surplus, lean season procurements were acceptable.)

m. If procurement in a country different from the country of distribution did not take place, why not?

n. Did the procurement take place as originally planned, or did the PVO have to identify substitutable commodities and/or new surplus markets as a result of changes in supply over the life of the project?

o. Does the project use vouchers? i. Cash-based vouchers? Coupon-like vouchers? ii. Were beneficiaries restricted to a limited number of commodity types? iii. How many vendors participated in the program? iv. How often were the vouchers distributed? v. What was the voucher redemption rate? vi. Were there any instances of counterfeiting or fraud?


vii. Did the PVO consistently monitor the market to ensure that there was adequate supply so as not to price out low-income consumers?

viii. Did the PVO attempt to monitor the quality of the commodities sold by the participating vendors?

p. Were local tastes in food considered? q. Were likely/possible market impacts considered? If so, how? r. Were there any challenges related to changes in commodity prices? (i.e. Did prices

increase/decrease over the life of the project and if so, what was the impact?) s. Were all of the commodities of developing country origin? (Note: Only commodities of

developing country origin were eligible for purchase under the pilot program.) t. If the USG was monetizing any commodities in the region, were any USG commodities

purchased under the project? u. Were there any delays in the procurement process itself, or in the delivery of the commodities? v. Were there any instances of supplier default? If so, what was the reason (i.e. a poor harvest, an

increase in commodity prices which enabled the vendor to sell to another buyer at a higher price, poor quality, etc.)?

w. Were there any market-distorting laws or regulations in place in the procurement country that had a negative impact on the procurement or the delivery of the commodities?

x. Were there any instances of interference by procurement country and/or recipient country governments specifically in response to the LRP project such as price-setting, taxation, etc. that impacted either the purchase price of the commodities or the delivery timeline?

y. Did the PVO staff feel that they paid a fair price for the commodities? Why or why not? z. Was LRP the appropriate response for the food security need identified in each project? (Was

the food security need due to a lack of availability or a lack of access?) aa. Were the commodities that were procured for the LRP projects sufficient to meet the needs in

every case? bb. Were there any instances in which in-kind food aid or a combination of in-kind and LRP would

have been a more appropriate response? cc. For emergency projects, if the food aid needs were to significantly increase, could LRP be scaled

up to respond to those needs? Could LRP be used to fill a larger gap in the food security needs, or would some in-kind assistance be required?

dd. Did the commodity or commodities procured locally or regionally fill a need that in-kind commodities could not? (Consider the benefits of processed, fortified commodities such as incaparina in Guatemala, and iron-fortified fish sauce in Cambodia for example, both of which were used for targeted nutritional interventions and are not available in-kind. There is the potential that culturally appropriate and locally processed fortified commodities could be cheaper than processed and fortified in-kind commodities for targeted nutritional interventions.)

ee. Did the commodities purchased through local or regional procurement complement an existing USG in-kind food aid program?

2. What has been your role with these pilot projects? 3. Have you visited the project procurement and distribution sites? Why or why not? If yes, what did

you learn from these visits that you might not have learned otherwise? 4. Are the procurement projects on track to achieve their objectives? Why or why not? What would you

do differently in the procurements? 5. Tell me about capacity building that has taken place with these activities. What has been learned, who

has learned it, and how is it used?


6. How does your project collect market price data?

a. What specific data are being collected? b. What is the range of data (commodity/market/frequency/zone) that your projects collect? c. What is the range of regional data that your projects collect? d. Do the data relate to wholesale, retail or farm-gate prices? If “retail”, is there a need to convert

price data to wholesale levels and how is this done? If collecting secondary data from other sources such as government databases, does the data track at the same price level?

e. How satisfied are you with the timeliness of the availability of data for decision making? f. What are the sources of these data? g. Is there a need to “clean” or “correct” the data? h. What comments do you have on the apparent quality of data from different sources, if available? i. To what extent do you graph, tabulate, summarize or analyze the data from a market impact

perspective? j. To what extent has analysis of past procurements informed decision-making on later

procurements under the project? k. What are the qualifications and task-specific training of the staff member responsible for the

collection of data and the analysis of these data? l. How many full-time-equivalent staff members work on the collection and analysis of market

data? m. Did the PVO have a methodology for determining a “reasonable” market price? (Note: The

Farm Bill requires that commodities be procured at “reasonable market prices.”)

7. What data beyond market prices for commodities to be procured do your projects collect? For instance, how do they obtain trucking cost data and other data necessary for import parity price (IPP) calculations? (Note: In some instances, trucking and storage costs will be built into the commodity prices in the contracts. Some contracts may have been structured to ensure just-in-time delivery due to insufficient storage at or near distribution sites.)

8. How was the IPP calculated? Was the PVO’s methodology for calculating the IPP appropriate? 9. Do you or your team work with U.S. in-kind food donations at any time? If so, what data do you

have on those projects?

10. How does your project assess the degree and duration of price rise that LRP activities will cause in markets? Do the approaches work? Are their limits on price rises acceptable?

11. Are there any other significant USG or other food aid activities being implemented in the same areas

as your project? Which? (i.e. in-kind food aid distributions, including monetization, host-country government procurements for food reserves or other safety-net programs, etc.) a. Are there any overlaps between the operating areas, beneficiaries, commodities, and objectives of

those activities and your LRP projects? If so, what are they and how will they affect your assessment of the LRP activities?

b. What is the size of the purchase market(s)? c. How well integrated is the purchase market? d. What effects are seen in the market as a result? How do you know how to attribute the effects? e. What is the market share of your LRP activities and that of these overlapping projects?


f. How well do those responsible for USG-funded procurements coordinate (a) their pre-procurement market assessments and (b) their resulting procurements?

g. What level of coordination exists between different actors responsible for food-aid procurement in the countries/regions where your projects operate?

12. Do you have evidence that your LRP activities have changed market structure or conduct (apart from

having caused rises)? For instance, do larger traders tend to supply a significantly higher proportion of the commodities procured?

13. How have you monitored the progress and achievements of your project(s)? What monitoring data do you collect?

[Burkina Faso, Zambia and Guatemala only:] What is the methodology of the impact evaluation underway in your country? Has the evaluation uncovered new information for you? 14. Why, when and how did procurement delays occur? Could they have been avoided? If so, how?

15. Is it difficult to find reliable local/regional suppliers for LRP activities? Do you use one (or a few)

suppliers with monopolies or oligopolies in the markets for specific commodities?

16. Tell me about any contractual problems with LRP. How easy is it to enforce contractual conditions about sack weights, quality, etc.? Do your projects require suppliers to purchase performance-guarantee bonds that they lose if they do not meet their contractual obligations?

17. Do your projects’ food shipments experience significant delays due to harassment of shipments at

road checkpoints or border posts, or due to pre-shipment bureaucratic corruption and/or delays? Do such problems affect LRP shipments more than in-kind shipments from the US?

18. Can you think of any practices or processes used in LRP that have been especially effective? 19. What kind of obstacles have you faced in carrying out your work on the LRP pilots? How have you

overcome these? 20. How well has your project addressed the quality of the procurements? Please explain.

a. Do you sample foodstuffs before purchase? If so, please describe your sampling methodology. b. For what do you test the samples taken? c. What set of standards and which thresholds do you apply to the tests? d. What happens to shipments that marginally fail a given test? e. Were there any significant challenges related to the commodity testing requirements such as the

lack of availability of appropriate testing services, delays in receiving results, etc.?

21. What unexpected LRP outcomes have you seen? 22. Do you think the LRP projects have adequately met the needs of your beneficiaries? Please explain.

a. What do the food aid recipients think about the LRP activities? b. What do other beneficiaries, such as farmers and vendors, think about the activities? c. What is a good measure of their level of satisfaction?


d. Has their food security/nutrition been demonstrably improved by LRP? 23. How could local and regional procurements be improved in terms of timeliness? What delays do

LRP shipments experience that in-kind shipments do not? What delays do in-kind shipments experience that LRP shipments do not?

24. Do you think resources have been used effectively and efficiently in the LRP pilots? Why or why not? Has the level of these resources been adequate?

25. Are there ways to reduce costs while maintaining quality? How?

26. What new skill sets do project staff need to perform LRP work that in-kind work does not require?

What skill sets do they need to process in-kind imports of food aid that they do not need for LRP?

27. How many more/less staff do you need per 10,000 metric tons of LRP food aid, compared with in-kind food aid?

28. How could LRP be improved, with regard to its effects on intended beneficiaries and markets?

29. Is there any evidence that the distribution of locally or regionally purchased commodities displaced

commercial sales?


3. Beneficiary Interview (Customized by Country; Cambodia IRD Project Version Included Here, Others Available upon Request)

Name of Respondent: ___________________________________

Please tell us about the food/voucher you received.

1. What exactly did you receive from the program? □ (1) Voucher for food □ (2) School meals □ (9) Other (specify)

2. Did the program require anything of you in exchange? (Probe for what)

[ IF SCHOOL MEALS/NOT VOUCHER]: How did you get the school meals?

3. Was this convenient for you? Why or why not?

4. Did you have any problems getting the school meals? [If yes, probe for the problems.]

5. Have you ever been unable to get the school meals, even when you wanted to do so? [If so, probe for how often and why.]

[IF VOUCHER/NOT MEALS]: How did you get the voucher?

6. Was this convenient for you? Why or why not?

7. Did you have any problems getting the voucher or the food? [If yes, probe for the problems.]

8. What restrictions did the voucher have, if any? [Probe for commodity/vendor restrictions.]

9. Have you ever been unable to redeem a voucher, even when you wanted to do so? [If so, probe for how often and why]


10. How often have you received school meals/vouchers from this program? [Circle meals/voucher] □ (1) Three or more times per week □ (2) Once or twice a week □ (3) Every other week □ (4) Less than twice a month □ (5) Varied schedule □ (9) Other [specify]:__________________

11. How long have you been receiving food/vouchers from this program? [Circle meals/voucher] □ (1) One month or less □ (2) One month to three months □ (3) Four months to six months □ (4) More than six months □ (5) Starting and stopping □ (9) Other [specify]:________________

12. Has the food/voucher helped you and your family? [Probe for how, or why not]

13. What was the quality of the food you received?

14. Was the food you received something you prefer to eat? [Probe for why/why not]

15. Is there anything you’d like to change about the program? [Probe for changes and why]

16. Please tell me how many adults live in your household, and how many children under age 15. Adults: __________ Children: _________

17. Of the total number of adults in the household [read the number], how many are physically able to work regularly (even if they do not have a regular source of work)? Adults who can work: ________

18. Of the total number of adults in the household [read the number], how many do have a regular source of work [either income-based or farming]? Adults who have a regular source of work: ________

19. Has your household suffered any problems that affected your ability to feed your family? These might include loss of life, health, or livelihood, or drought or other problem. [If so, probe for problem, when, coping strategies.]

Now I have some questions about your welfare.

20. Would you say you are not poor, somewhat poor, or very poor? □ (1) Not poor □ (2) Somewhat poor □ (3) Very poor □ (9) Other response [specify]:___


21. In the past one month, were you ever worried about whether your family would get enough to eat? □ (1) Yes □ (2) No □ (9) Other response [specify]:____

22. Please tell me if any of the following has occurred in your family in the past one month due to lack of resources to get food. [Read all possible cases to respondent, tick the box if they say “Yes” to any.] □ (1) You or other household members had to eat food you do not prefer □ (2) You or other household members had to eat less than you felt was necessary □ (3) You or other household members had no food in the house □ (4) You or other household members went to sleep hungry □ (5) You or other household members went an entire 24 hours without eating

23. [If “yes” to any of the statements in previous question]: For approximately how many months has this been happening to you and your household? [Enter # of months.] ___ (1) You or other household members had to eat food you do not prefer ___ (2) You or other household members had to eat less than you felt was necessary ___ (3) You or other household members had no food in the house ___ (4) You or other household members went to sleep hungry ___ (5) You or other household members went an entire 24 hours without eating

24. What effects, if any, has the food/voucher program had on your household expenses?

25. In the past month, have the prices for food in your market(s) changed? [Probe for direction of change]

26. [If yes]: What did you do when prices changed?

27. In the past month, have the foods you want to buy usually been available in your market(s)?


4. Non-Beneficiary Interview (as above; Zambia Land O’Lakes Project Version Included Here)

NON BENEFICIARY INTERVIEW - Zambia

Name of Respondent: ______________________________________________________

1. Do you know of any food programs in your community? If so, what are they?

2. Has your household suffered any problems that affected your ability to feed your family? These might include loss of life, health, or livelihood, or drought or other problem. [If so, probe for problem, when, coping strategies.]

3. In the past one month, were you ever worried about whether your family would get enough to eat?

4. In the past month, have the prices for food in your market(s) changed? [Probe for direction of change]

5. [If yes]: What did you do when prices changed?

6. In the past month, have the foods you want to buy usually been available in your market(s)?

7. Have you noticed anyone re-selling any food aid goods here?


5. Wholesale Vendor Survey Instrument

1. General information about the vendor

a) What is the size of your business (employees, how many branches/where, and estimate of sales)

b) What type of client do you usually do business with?

c) How many contracts do you usually get a year, on average?

d) How much are you usually required to supply per purchase order (get tonnage if possible)?

e) Usually, when supplying donor programs: where do you obtain the food (locally, regionally)?

f) Usually, when supplying donor programs: where is it delivered (how far from the purchase site)?

g) Usually, when supplying donor programs: how is the food delivered?

h) If you have other clients (than donors), how different is it to purchase for these other clients?

i) What are the obstacles/constraints you encounter most in the course of your business?

2. Project-specific information

2.a. Contractual matters:

a) How were you made aware that [insert PVO name] was looking for a vendor to purchase food?

b) How large was the purchase?

c) Was it a one-time purchase or was it divided into several tranches?

d) What provisions did the contract have other than the amount and type of food to be purchased (e.g. quality, place and speed of delivery, etc.)?

e) What would you have liked to have done differently in your business with [insert PVO name] that would allow you to fulfill your contract better (time and quality) (e.g. have more time to fulfill the contract, have clearer instructions, different payment methods/schedule, etc.)?

2.b. Logistics:

a) Where was the purchase made (how far from delivery point)?

b) Did you purchase one, or several commodities?

c) Once gathered in one place and ready for transportation, how was the food transported? How far (kilometers)?

d) How long did it take from the moment the commodity shipment was ready for shipment to the time of delivery?

e) How long did transportation take to the delivery point?

f) Did the commodity transported have any logistical requirement(s) specific to the type commodity?

g) Were there any constraints you encountered during transport?

h) What were these constraints due to?

i) What additional costs did those constraints generate?


j) Were you able to address these constraints and how?

k) What would you have liked to have done differently in terms of how you handled the delivery?

l) Quality testing – what was this process like?

m) How long did the quality testing process take?

n) How is this quality testing process the same or different from that of your other buyers?

3. Market information

a) Was the purchase made in a time when food prices were higher? Lower? (Sensitive question)

b) Around the time of purchase (purchase +/- 4 months), have market prices for these commodities fluctuated a lot and if so, do you know what caused those fluctuations?

c) Do you have a sense of how much, compared to the total supply available, your aggregated purchases for this contract represented?

d) In your opinion, did the purchases for this contract affect food supply, demand and prices where they were made?

4. Buying process

a) How does the buying process (from producers) work in cases like the one related to your contract (description of intermediaries, acquiring pricing information, description of potential networks, price negotiations, transport, packaging)?

b) How easy/difficult was it for you to obtain the totality of the supplies from producers?

c) What constraints did you encounter?

d) What were these constraints due to?

e) How did you address them?

f) How long did it take from contract signature to having all the food ready for delivery?

g) Ideally, what would be different in terms of the market you face in cases such as this one?

5. Increase in business

a) How have you benefited from your business with [insert PVO name]?

b) Reputation and visibility

c) Quality improvements

d) Capacity building

e) Other


6. Vendor Questionnaire – Retail (as above; Uganda World Vision Project Version Included Here)

Local (Retail) Vendor interview - Uganda

1. Have prices changed since December, 2010? (In general)

Price changes for maize grain? (How much, in what direction)

Price changes for peas and beans?

Price changes for vegetable oil?

2. Have supplies for these products changed, since December 2010?

3. Have there been any changes in the market since December 2010? (Probe for shock, weather issue, world prices, other market changes.)

4. Have there been any changes in the quality of the products you sell since December 2010?

5. How was your experience with the World Vision voucher program? [Omit if never a vendor for WV program. Topics to discuss: paperwork, payment, value for vouchers.]


ANNEX VI. METHODOLOGIES


CONTENTS

REVISED TIME AND COST COMPARISON METHODOLOGY .................................. 242

MARKET PRICE IMPACT METHODOLOGY ................................................................. 246


Revised Time and Cost Comparison Methodology

1. Time Timeline examined

Table 1 below shows steps in Local/Regional Procurement and in-kind procurement that are functionally equivalent and that we will use to compare timeliness. We divide the LRP approach into two columns (food distributions and vouchers) corresponding to two modalities of LRP. In-kind procurement is divided into shipments using a Through Bill of Lading (TBL) and those not using a TBL. Prepositioning is shown in Table 2.

TABLE 1: STEPS IN IN-KIND AND LOCAL/REGIONAL PROCUREMENT THAT ARE FUNCTIONALLY EQUIVALENT AND COMPARABLE

LRP In-Kind Procurement

Food Distributions Vouchers Using Not Using

Through Bill of Lading Through Bill of Lading

Date of issuance of hard or soft tender by participant.

Date of issuance of soft tender by participant, or search for suppliers by participant.

Date of issuance of a solicitation for commodity and freight by FSA/KCCO.

Date of issuance of a solicitation for commodity and freight by FSA /KCCO.

Contract signature. Vendor certification or announcement of voucher fair.

Freight contract (“booking note” for processed/packaged commodities, or “charter party” for bulk commodities) awarded to vessel owner and Purchase contract issued to commodity vendor. If there are two different dates for these contracts, we will take the latter date, as both are required for shipping to occur. Where booking notes or charter party documents are not available, or dates are not included, the freight contract award date will be used for this midpoint.

Commodities arrive at delivery endpoint; transport arranged by supplier or participant. In case of inland transport covered by participant, this time will be included here.

No equivalent step; use voucher fairs/voucher-based purchases for comparison.

Commodities arrive at delivery endpoint (Inland transport is included in these contracts, for transport from discharge port through point of entry in the recipient country.)

Commodities arrive at delivery endpoint (In these cases, inland freight times must be captured for Title II shipments where participants are responsible for inland transport.)

All steps in the table are functionally comparable across modalities and will be used as variables to calculate corresponding durations for comparison. In all cases, the starting point in our timeline is the moment when tenders are issued or any search mechanism for suppliers/vendors is initiated. This goes back far enough in the LRP/in-kind process to be comprehensive while avoiding non-comparable administrative steps; it illustrates well how easy (hard)/fast (slow) it is to find suppliers; and it allows for the inclusion of processes that differ from the more traditional tendering methods, especially when dealing with small traders or vendors. The end point of our timelines are the dates when the commodities arrive at the participant’s delivery endpoint, i.e. the location (typically, a warehouse) usually close to or within the distribution zone from which commodities will be dispatched to be distributed. Because in voucher approaches this step does


not exist, we propose to use the dates of voucher-based purchases or voucher fairs, with implications outlined below.

Food distributions (LRP): We use as a starting point the date of issuance of a hard (competitive) or soft (non-competitive or limited) tender by the participant. The next variable we use is the signature of the contract, corresponding to the point where the participant has selected one or several suppliers with whom an understanding has been agreed upon to purchase commodities. The next and last step we use is the arrival of the commodities at the participant’s delivery endpoint, i.e. the location (typically, a warehouse) usually close to or within the distribution zone from which commodities will be dispatched to be distributed. This end date must include quality testing; if not, the end date of quality testing and any remedial actions as a result of testing will be the end date.

Vouchers (LRP): We use as a starting point the date of issuance of a soft (non-competitive) tender by the participant, or when the project-specific search mechanism for suppliers/vendors is initiated. The next point in the analysis is the announcement of a voucher fair or, in absence of a fair, the date of certification of selected vendors. This corresponds to the point where there is agreement with several vendors that the participant’s vouchers will be redeemable after exchanges with beneficiaries have taken place according to pre-specifications (i.e. quantities, commodity types, etc.). Contrary to the direct distribution modality, in the voucher case there is no arrival of the commodities at the participant’s delivery endpoint. Essentially, the delivery endpoint is skipped and the beneficiaries directly access the commodities through approved vendors. We will therefore use as the final milestone, voucher fairs or, in the absence of fairs, the start date of voucher-based purchases. This will come from participants’ accounting datasets indicating the dates recorded on vendors’ reimbursement paperwork. It effectively establishes the time when beneficiaries are able to access commodities through the use of the voucher system. Our study will examine if, and to what extent, this might benefit the voucher modalities in terms of time performance.

LRP voucher programs were iterative, in most cases, in that a “round” of vouchers included several steps, and multiple (two or more) “rounds” of vouchers were distributed in each project site, over the life of the project. The first such distribution “round” would have involved the administrative steps to prepare for distribution (including identifying and training of vendors, printing of vouchers, etc., as listed above), followed by voucher distribution, redemption of the vouchers by beneficiaries, and a closing out of accounts with vendors. Later “rounds” would have left out the first time- and cost-intensive step; therefore, for purposes of comparison, we will examine, compare and discuss the difference in timelines between initial and subsequent rounds. Time for quality testing of vendor supplies will be considered as well; however, participant staff have indicated that quality testing was completed before distribution of vouchers, and so that timeline would be included in the voucher timeline as described above.

Shipments using TBL: The starting point is the date of a solicitation for commodity and freight by FSA/KCCO. The next point is the award of a freight contract to the vessel owner and a purchase contract to the commodity vendor, corresponding to (respectively) when the booking note or charter party has been signed. Both awards need to have been made for this step to be considered complete. The last is the arrival of the commodities at the participant’s delivery endpoint. With shipments using a Through Bill of Lading the participant does not need to take possession of the commodities at the port of entry; instead they can be transported all the way to the delivery endpoint. Our study will explore whether or not this results in gains of delivery time.

Shipments not using TBL: The starting point and the next point are the same as with shipments using a TBL. The exception for shipments not using a TBL is that the participant usually takes possession of the commodities at the port of entry and arranges for inland shipment to the delivery endpoint. In these cases, where inland transport is the responsibility of the participant, the elapsed time and costs for this step will be included. This is in parallel to LRP projects where inland transport is used to bring commodities to the recipient country.


Table 2 below shows the steps in the procurement process for in-kind and prepositioned food aid110 that are functionally equivalent and that we will use to compare timeliness. While in the case of prepositioning, arranging for commodity purchases and transportation from the United States are steps that have been completed prior to the expression of an explicit need for commodities for a particular emergency or development project, there is still a process taking place for prepositioned commodities (or, in some cases, commodities that are en route) to be released and transported to a particular destination. We ask for discussions with the relevant USG personnel to allow us to become better acquainted with these steps and make pertinent choices for our comparisons.

TABLE 2: STEPS IN IN-KIND PROCUREMENT AND PRE-POSITIONING THAT ARE FUNCTIONALLY EQUIVALENT AND COMPARABLE

In-Kind Prepositioning111

Through Bill of Lading Other Bill of Lading



Contacted the relevant USG personnel to obtain specific information to identify these steps.

Freight contract awarded to vessel owner and Purchase contract issued to commodity vendor.

Freight contract awarded to vessel owner and Purchase contract issued to commodity vendor.

Contacted the relevant USG personnel to obtain specific information to identify these steps.

Commodities arrive at delivery endpoint; transport arranged by ocean freight carrier.

Commodities arrive at delivery endpoint; transport arranged by participant.

Commodities arrive at delivery endpoint – may be through diverted shipments, or from pre-positioned warehouses.

Time analysis

Calculation of the duration for each step will start on the date of the functions in the first cell for each modality, in both Tables 1 and 2. End dates are from the last cell for each modality. In-kind and pre-positioning shipments with comparable delivery dates to those of matched LRP shipments will be used as comparisons.112 Comparability is defined as within three months prior to or after LRP delivery date.

For each step, we will calculate a Standard Deviation, eliminate durations outside ±1.5 SD, and calculate the average time taken per project, per commodity per project, and per approach. Although we will discard the outliers from the calculation of the averages, we will examine them and draw inferences and lessons where possible and relevant. Where a basket of commodities is to be distributed, we will look at the longest time for any commodity in that basket to be distributed (since the basket of commodities cannot be distributed until all of the commodities are at hand). We will also calculate an average timeline per ton of the basket, calculated as an average of the times, weighted by the tonnage proportion of each commodity.

Total response time for emergency projects is a special category of comparisons that will look at a subsample of projects designed to meet needs during humanitarian crises. Those LRP projects intended to provide

110An additional modality was considered for the comparison: the Forward Purchasing Facility currently being tested by WFP. However, as these operations are very new, and insufficient data exist for purchases under this Facility, these types of purchases were excluded from our analyses. 111 In May of 2012, MSI received notice from USDA that the evaluation team was not to include prepositioning in this report, as data would not be made available. 112Matching involves commodity type – cereals to cereals, pulses to pulses, oils to oils. In those cases without acceptable matches from among in-kind shipments, such as for processed or fortified foods like pinollo and incaparina, MSI will seek shipments of CSB or other fortified foods and weigh the time and cost per kilocalorie between the modalities.


emergency food assistance will receive this additional analysis, and will be compared, when possible, to any pre-positioned commodity shipments made for emergency purposes.113

An additional line of inquiry under timeliness comparisons is a comparison of time to source commodities. In addition to the quantitative comparison between cells one and two per modality, our qualitative data from LRP site visits will augment our understanding of sourcing issues encountered by LRP participants. It should be noted that this is especially relevant for emergency projects; in development projects, longer time periods for sourcing may be related to direct project development goals. These issues will be explored in the timeliness analysis across visited projects.

2. Costs Costs included

We propose to examine costs that are comparable across procurement types and participants, such as (in total USD and in USD per metric ton) (1) the cost of the commodities, (2) the aggregated cost of transportation, and (3) handling and storage costs, including costs for commodity quality and safety testing.114 To the extent our data permit it, we will try to elicit differences for subcategories of costs (e.g. fumigation costs; inland transportation costs) in the different procurement approaches/modalities. We leave activity and administrative costs, which are structured differently across participants and not directly tied to the approach used, out of the comparison.

Just as voucher approaches exclude a significant step (the arrival of the commodities at the delivery endpoint) in our proposed timeline, they also do not require the incurrence of storage or handling costs – therefore creating a positive a priori bias in favor of cost comparisons between voucher and non-voucher approaches.

However, voucher approaches do require administrative costs that are specific to them and avoided under other modalities, so including those costs will allow us to correct the bias in an attempt at a fair representation of costs. The voucher costs we propose to include are printing costs, voucher distribution costs, as well as costs related to “training” vendors and beneficiaries in the voucher process. We exclude monitoring or personnel training costs, as those are not particular to voucher programs.

Pre-positioned commodities and shipments may include unique cost categories. Because the commodities taken from pre-positioned inventory are requested on short notice at times, vessels carrying commodities for pre-positioning are sometimes diverted before they even arrive at the pre-positioned warehouse. Other times, commodities are in the pre-positioned warehouse and shipped directly to project sites from there. Total costs for pre-positioning must also include warehouse maintenance costs such as rent and security (prorated), and additional costs from diversion to unplanned discharge ports.

Cost analysis

We will calculate the costs of each cost category (i.e. transportation, warehousing, handling, etc.) per project and of each category per commodity per project. We will also calculate the costs of each category per approach/modality (e.g. the average cost of transportation for in-kind food aid and LRP). Analyses will examine differences in costs within categories (such as examining the data by distribution type) and findings will be discussed in the report text.

113 MSI and USDA requested data on pre-positioned commodities, but this data was not made available to the MSI evaluation team. 114For the in-kind shipments, this is done prior to transoceanic shipment; for LRP, it is done in-country.


Market Price Impact Methodology

Introduction Explain:

A. Why procurement should increase demand and thus price in the procurement zone. B. That there’s an issue of detectability: it’s possible that the price increase may be too small to differentiate

from background statistical noise. C. That while a food-aid distribution may be limited to a small area and therefore have a concentrated effect

on depressing local market prices, a food-aid procurement of the same size may take place over a much larger area leading to a muted price increase in markets across that greater area and thus to greater difficulty in detection of that price increase.

Detection of the demand-side impact due to food-aid distributions

We will start with a brief examination of the price impact of food-aid distributions because this should be an easier case to make (for the reason C. above), with a discussion of targeting of food aid and its associated income and substitution effects.

Evidence for detectable price falls due to food-aid distribution:

A. There exist several examples, with greater or lesser explanatory force, of testimony from knowledgeable informants reinforcing the hypothesis that distribution of a given commodity depresses its local price. In particular, we have persuasive oral evidence of the price-lowering effect of food-distributions in Niger.

B. We assume that there is probably also evidence from inspection of price series of given food-aid commodities in local markets for that commodity.

Conversely, we will then use these approaches and more formal analysis to look for supply-side price spikes caused by local or regional procurement.

Key informants’ statements about market impact

Firstly, we will report the results of interviews with key informants in procurement zones supporting a price-spike effect due to an LRP procurement.

Choosing the most likely candidates for finding price rises due to procurement

Secondly, we will look at the evidence for an impact due to procurement, focusing our attention on the data sets judged most likely to show this evidence. In order to do so, we will rank-order the price data sets by likelihood of detecting a price spike, using several criteria:

A. Evaluation of data quality, favoring evaluation of data sets meeting the following criteria: 1. Primary data from a given procurement market collected as part of an LRP project, rather than

secondary data (unless the secondary data happen, by chance, to be exactly what is needed). 2. Procurement-market data alongside data for control markets (i.e. similar markets subject to the same

exogenous forces, but beyond the direct impact of food-aid disbursements in the immediate catchment area), so that we can directly compare ‘with’ and ‘without’.

3. Price data series sufficiently long to allow analysis from three months before the award of an LRP contract until three months after the last commodity delivery to the PVO/WFP (or three months after beneficiaries redeem vouchers). The ± 3-month period will accommodate the possibilities of


market actors anticipating purchases after news of the procurement has reached the market and also any lags in market reaction to the impact of extra purchases having taken place in the market.

4. Price series data are complete, rather than patchy. 5. The frequency of price series data is weekly or more frequent, rather than monthly or less frequent. 6. Price data pertain to exactly the commodity procured, not a substitute. 7. Price data have been collected for the market or zone where the procurement took place, rather than

elsewhere.

B. Exogenous variables:

Note the degree to which prevailing exogenous variables other than LRP procurement are relevant, e.g.:

1. Seasonal effects esp. agricultural calendar (and particularly, harvest period). 2. Inter-annual effects, e.g. poorly spaced rainfall or a reduction in cumulative rainfall over the rainy

season, relative to the seasonal norms. 3. The timing of other simultaneous significant procurements (but see C. below). 4. The timing and scale of changes in government policy, e.g. export bans. 5. Civil unrest and security issues.

C. Ideally econometric analysis would explicitly control for these factors. Absent this capability, to the extent that it is difficult to attribute price variation to exogenous variables, the data set will fall down the ranked order LRP procurement-distribution overlap.

If the LRP procurement took place in the same zone as the corresponding LRP distribution, do we have timing of the sequence of the possibly overlapping sequence of purchases and distributions, with their corresponding geography? If LRP distribution overlaps geographically with LRP procurement, the two market impacts would partially cancel each other out but, to the extent that it is unclear if the net price profile would allow us to disentangle these two effects, the procurement will fall down the ranked order.

Examination of likely causes for price increases induced by LRP procurement

Working initially with the most promising of the rank-ordered data sets, search for evidence of price spikes due to LRP procurement. Evaluate the evidence for price spikes in terms of:

A. Visual inspection of the price data, taking into account B. and C. above, i.e.:

1. If procurement and distribution take place in close mutual proximity, look for data to allow us to disaggregate the effects of each.

2. Account for seasonality, inter-annual variations and other exogenous variables. 3. Note spikes for which LRP procurement seems likely in terms of plausible timing, with no other

superior explanation.

B. For markets (or procurement zones) for which the typical sales volume of a given commodity is known or, less satisfactorily, can be estimated:

1. Null hypothesis, H0: LRP procurements of the commodity caused no measurable price spikes because relatively small quantities were purchased over wide areas.

2. State and explain the equation for the elasticity of supply (Ƞs):

Ƞs = (ΔQ/Q) / (ΔP/P) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (1)


Where ΔQ is the change in supply to the market (in this case due to LRP procurement), Q is the market throughput, P is the measured price level, and ΔP is the change in price possibly due to LRP procurement.

3. Note the formal limitations due to omitted variables on the use of equation (1) to characterize the relations between the quantity supplied to the market and the resulting price level.

4. Determine the plausible range of elasticities from the supply-response literature for agriculture. 5. Establish from the price data a level that ΔP would have to exceed to be distinguishable from

background statistical noise during the ± 3-month period of interest. 6. Infer the market throughput, Q, from equation (1) and compare the estimates corresponding to

plausible upper and lower values of the elasticity of supply with the known or estimated value. 7. If an independent estimate of the value of Q differs significantly from the value generated via

equation (1), review where errors may lie and reconcile the evidence. 8. Note where the evidence from visual inspection and/or operationalization of the elasticity equation

suggests an LRP-generated price spike.

If none of the most promising data sets shows evidence of LRP-induced price increases, we accept the null hypothesis and accept that there is no justification to examine the other rank-ordered data sets in which it is even less likely that we will be able to discern any similar impact.

Conversely, if one or more of the most promising rank-ordered data sets shows evidence of LRP-induced price increases:

1. Note the scale and duration of those increases, as well any lags with respect to procurements. 2. Inspect a second tranche of likely data sets for similar impacts. 3. Continue until inspection reveals a tranche with no further LRP effects. 4. Exclude from further examination those data sets not amenable to analysis by either visual inspection

or elasticity-based analysis.

Conclusions & recommendations

Note the extent and strength of any evidence for price spikes due to LRP.

Draw conclusions about the usefulness of the data for these analyses. Cross reference to other studies, e.g., GAO reports that comment on the limited quality of the data for the purposes of such analysis.

If LRP hasn’t had any measurable effect, estimate the level of ΔQ that would generate a detectable ΔP. Compare this level with the actual level of LRP procurements.

Discuss the limitations and utility of the analytical approaches used.

USDA LOCAL AND REGIONAL FOOD AID PROCUREMENT PILOT PROJECT · “USDA Local and Regional Food Aid...

Documents

Transcript of USDA LOCAL AND REGIONAL FOOD AID PROCUREMENT PILOT PROJECT · “USDA Local and Regional Food Aid...